Data Analysis
26820228
10/08/2021
Information
Please refer to ‘Data Cleaning’ script prior to accessing this script.
Setup
knitr::opts_chunk$set(echo = TRUE)
require("knitr")## Loading required package: knitropts_knit$set(root.dir = "~/Library/Mobile Documents/com~apple~CloudDocs/Documents/Uni/Masters/Empirical Project/Code/Empirical_Project")
# turn off scientific notation
options(scipen = 999)
## colour palette
# e7b553
# c45150
# a24b6f
# 824372
# 603863
# 382c46
# 403250Load Libraries
library("ggplot2") # for figures
library("psych") # for Cronbach's alpha, for describe function##
## Attaching package: 'psych'## The following objects are masked from 'package:ggplot2':
##
## %+%, alphalibrary("ppcor") # for partial correlation p-values## Loading required package: MASSlibrary("dplyr") # for mutate function##
## Attaching package: 'dplyr'## The following object is masked from 'package:MASS':
##
## select## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary("ggpubr") # for qq-plots
library("GGally") # for scatterplot matrix ## Registered S3 method overwritten by 'GGally':
## method from
## +.gg ggplot2library("effsize") # for calculation of effect size##
## Attaching package: 'effsize'## The following object is masked from 'package:psych':
##
## cohen.dlibrary("pwr") # for power calculation
library("performance") # for assessing robustness of model
library("effsize") # for eta squared
library("reshape2") # for transforming data from wide to long format
library("tidyverse") # for data cleaning ## ── Attaching packages ─────────────────────────────────────── tidyverse 1.3.1 ──## ✓ tibble 3.1.3 ✓ purrr 0.3.4
## ✓ tidyr 1.1.3 ✓ stringr 1.4.0
## ✓ readr 2.0.0 ✓ forcats 0.5.1## ── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
## x psych::%+%() masks ggplot2::%+%()
## x psych::alpha() masks ggplot2::alpha()
## x dplyr::filter() masks stats::filter()
## x dplyr::lag() masks stats::lag()
## x dplyr::select() masks MASS::select()library("rstatix") # for ANOVA and ANCOVA##
## Attaching package: 'rstatix'## The following object is masked from 'package:MASS':
##
## select## The following object is masked from 'package:stats':
##
## filterlibrary("gridExtra") # for grid.arrange function##
## Attaching package: 'gridExtra'## The following object is masked from 'package:dplyr':
##
## combinelibrary("car") # for levene's test## Loading required package: carData##
## Attaching package: 'car'## The following object is masked from 'package:purrr':
##
## some## The following object is masked from 'package:dplyr':
##
## recode## The following object is masked from 'package:psych':
##
## logitlibrary("emmeans") # to obtain estimated marginal means##
## Attaching package: 'emmeans'## The following object is masked from 'package:GGally':
##
## pigsSet Working Directory
# please change this to your own working directory path
setwd("~/Library/Mobile Documents/com~apple~CloudDocs/Documents/Uni/Masters/Empirical Project/Code/Empirical_Project")Read in Data and Save Data to an Object
# please change this to however you have stored the data file
# reading in dataframe 2, as this is the one with exclusion of n = 5
df <- read.csv(file = "data/cleaned/dataframe_2.csv", header = TRUE, na.strings = "NA")Change Variable Classifications
# change variable classifications to meet requirements for later analyses
# ensure IVs and categorical variables are factor variables
# and DVs or continuous variables are numeric variables
# participant id and demographics
df$id <- factor(df$id)
df$age <- as.numeric(df$age)
df$sex <- factor(df$sex)
df$ethnicity <- factor(df$ethnicity)
df$sexual_orientation <- factor(df$sexual_orientation)
# fixation count DVs
df$acq_csp_fix_count <- as.numeric(df$acq_csp_fix_count)
df$acq_csm_fix_count <- as.numeric(df$acq_csm_fix_count)
df$ext_csp_fix_count <- as.numeric(df$ext_csp_fix_count)
df$ext_csm_fix_count <- as.numeric(df$ext_csm_fix_count)
df$e_ext_csp_fix_count <- as.numeric(df$e_ext_csp_fix_count)
df$l_ext_csp_fix_count <- as.numeric(df$l_ext_csp_fix_count)
df$e_ext_csm_fix_count <- as.numeric(df$e_ext_csm_fix_count)
df$l_ext_csm_fix_count <- as.numeric(df$l_ext_csm_fix_count)
# fixation duration DVs
df$acq_csp_fix_duration <- as.numeric(df$acq_csp_fix_duration)
df$acq_csm_fix_duration <- as.numeric(df$acq_csm_fix_duration)
df$ext_csp_fix_duration <- as.numeric(df$ext_csp_fix_duration)
df$ext_csm_fix_duration <- as.numeric(df$ext_csm_fix_duration)
df$e_ext_csp_fix_duration <- as.numeric(df$e_ext_csp_fix_duration)
df$l_ext_csp_fix_duration <- as.numeric(df$l_ext_csp_fix_duration)
df$e_ext_csm_fix_duration <- as.numeric(df$e_ext_csm_fix_duration)
df$l_ext_csm_fix_duration <- as.numeric(df$l_ext_csm_fix_duration)
# saccade amplitude DVs
df$acq_csp_sacc_amplitude <- as.numeric(df$acq_csp_sacc_amplitude)
df$acq_csm_sacc_amplitude <- as.numeric(df$acq_csm_sacc_amplitude)
df$ext_csp_sacc_amplitude <- as.numeric(df$ext_csp_sacc_amplitude)
df$ext_csm_sacc_amplitude <- as.numeric(df$ext_csm_sacc_amplitude)
df$e_ext_csp_sacc_amplitude <- as.numeric(df$e_ext_csp_sacc_amplitude)
df$l_ext_csp_sacc_amplitude <- as.numeric(df$l_ext_csp_sacc_amplitude)
df$e_ext_csm_sacc_amplitude <- as.numeric(df$e_ext_csm_sacc_amplitude)
df$l_ext_csm_sacc_amplitude <- as.numeric(df$l_ext_csm_sacc_amplitude)Internal Consistency of IUS and STICSA
## IUS total
# compute & extract alpha value and save as an object
alpha_ius <- psych::alpha(df[, c("ius_1", "ius_2", "ius_3", "ius_4",
"ius_5", "ius_6", "ius_7", "ius_8",
"ius_9", "ius_10", "ius_11", "ius_12",
"ius_13", "ius_14", "ius_15", "ius_16",
"ius_17", "ius_18", "ius_19", "ius_20",
"ius_21", "ius_22", "ius_23", "ius_24",
"ius_25", "ius_26", "ius_27")])$total[1]
## STICSA total
# compute & extract alpha value and save as an object
alpha_sticsa <- psych::alpha(df[, c("sticsa_1", "sticsa_2", "sticsa_3", "sticsa_4",
"sticsa_5", "sticsa_6", "sticsa_7", "sticsa_8",
"sticsa_9", "sticsa_10", "sticsa_11", "sticsa_12",
"sticsa_13", "sticsa_14", "sticsa_15", "sticsa_16",
"sticsa_17", "sticsa_18", "sticsa_19", "sticsa_20",
"sticsa_21")])$total[1]
# create table of both Crobach's alpha values
cronbachs_alpha_questionnaires <- rbind(alpha_ius, alpha_sticsa)
# clean up row and column names for easier interpretation
rownames(cronbachs_alpha_questionnaires) <- c("IUS-27", "STICSA")
colnames(cronbachs_alpha_questionnaires) <- "Cronbach's Alpha"
# obtain Cronbach's alpha table
cronbachs_alpha_questionnaires## Cronbach's Alpha
## IUS-27 0.9496736
## STICSA 0.8766597Compute Questionnaire Totals
#### IUS total
# all items, no reverse scoring
df$ius_total <- as.numeric(df$ius_1 + df$ius_2 + df$ius_3 + df$ius_4 + df$ius_5 + df$ius_6 +
df$ius_7 + df$ius_8 + df$ius_9 + df$ius_10 + df$ius_11 + df$ius_12 + df$ius_13 +
df$ius_14 + df$ius_15 + df$ius_16 + df$ius_17 + df$ius_18 + df$ius_19 + df$ius_20 +
df$ius_21 + df$ius_22 + df$ius_23 + df$ius_24 + df$ius_25 + df$ius_26 + df$ius_27)
#### STICSA total
# all items, no reverse scoring
df$sticsa_total <- as.numeric(df$sticsa_1 + df$sticsa_2 + df$sticsa_3 +
df$sticsa_4 + df$sticsa_5 + df$sticsa_6 +
df$sticsa_7 + df$sticsa_8 + df$sticsa_9 +
df$sticsa_10 + df$sticsa_11 + df$sticsa_12 +
df$sticsa_13 + df$sticsa_14 + df$sticsa_15 +
df$sticsa_16 + df$sticsa_17 + df$sticsa_18 +
df$sticsa_19 + df$sticsa_20 + df$sticsa_21)Create High / Low IU Classifications
# compute variable classifying participants as high/ low IU on basis of median split,
# and store as factor
df$iu_group <- factor(ifelse(df$ius_total >= 65, 1, -1))
# high IU = 1
# low IU = -1Check Distribution and Range to Identify Extreme Scores and Potential Data Errors in Questionnaires
For IUS 27 Total in Both Groups
# possible total scores for the IUS range from 27-135
########################## check distributions
hist_ius_total <- df %>%
ggplot(aes(ius_total, fill = iu_group)) +
geom_histogram(binwidth = 5, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(20, 140, 10)) +
labs(x = "IUS-27 Total", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram of IUS-27 Scores") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_ius_total
# save plot to file
ggsave(filename = "graphs/histograms/hist_ius_total.png",
plot = hist_ius_total,
width = 20,
height = 10,
dpi = 300,
units = "cm")
########################## check ranges
range_ius_total <- by(df$ius_total, df$iu_group, range)
range_ius_total ## df$iu_group: -1
## [1] 32 64
## ------------------------------------------------------------
## df$iu_group: 1
## [1] 65 125# for high IU: 65-125
# for low IU: 32-64
##### overall: all scores are in range of possible scores, no errors apparentFor STICSA Total in Both Groups
# possible total scores for the STICSA range from 21-84
########################## check distributions
hist_sticsa_total <- df %>%
ggplot(aes(sticsa_total, fill = iu_group)) +
geom_histogram(binwidth = 5, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(20, 90, 10)) +
labs(x = "STICSA Total", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram of STICSA Scores") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_sticsa_total
# save plot to file
ggsave(filename = "graphs/histograms/hist_sticsa_total.png",
plot = hist_sticsa_total,
width = 20,
height = 10,
dpi = 300,
units = "cm")
########################## check ranges
range_sticsa_total <- by(df$sticsa_total, df$iu_group, range)
range_sticsa_total ## df$iu_group: -1
## [1] 22 57
## ------------------------------------------------------------
## df$iu_group: 1
## [1] 30 69# for high IU: 30-69
# for low IU: 22-57
##### overall: all scores are in range of possible scores, no errors apparentCompute Demographics
#### for age
# for all participants
all_age_table <-
describe(df[, "age"])
# for high IU
high_iu_age_table <-
describe(df[df$iu_group =="1", "age"])
# for low IU
low_iu_age_table <-
describe(df[df$iu_group =="-1", "age"])
# combine in a table
age_table <- rbind(all_age_table, high_iu_age_table, low_iu_age_table)
# re-name rows for easier interpretation
rownames(age_table) <- c("Age (All Participants","Age (High IU Group)",
"Age (Low IU Group)")
### for sex
sex_table <- xtabs(~ iu_group + sex, data = df)
sex_table <- prop.table(sex_table) %>%
round(digits = 4) * 100
rownames(sex_table) <- c("Low IU", "High IU")
sex_table## sex
## iu_group Female Male
## Low IU 26.28 24.82
## High IU 34.31 14.60### for sexual orientation
sexual_orientation_table <- xtabs(~ iu_group + sexual_orientation, data = df)
sexual_orientation_table <- prop.table(sexual_orientation_table) %>%
round(digits = 4) * 100
rownames(sexual_orientation_table) <- c("Low IU", "High IU")
sexual_orientation_table## sexual_orientation
## iu_group Heterosexual Sexual Minority
## Low IU 42.15 7.44
## High IU 42.98 7.44### for ethnicity
ethnicity_table <- xtabs(~ iu_group + ethnicity, data = df)
ethnicity_table <- prop.table(ethnicity_table) %>%
round(digits = 4) * 100
rownames(ethnicity_table) <- c("Low IU", "High IU")
ethnicity_table## ethnicity
## iu_group Asian Black Middle Eastern/ Arab Mixed White
## Low IU 7.26 1.61 2.42 0.81 37.90
## High IU 16.13 0.00 0.81 0.81 32.26#### write each to csv
# age
write.csv(age_table, file = "tables/demographics/age_table.csv",
row.names = TRUE)
# ethnicity
write.csv(ethnicity_table, file = "tables/demographics/ethnicity_table.csv",
row.names = TRUE)
# sex
write.csv(sex_table, file = "tables/demographics/sex_table.csv",
row.names = TRUE)
# sexual orientation
write.csv(sexual_orientation_table, file = "tables/demographics/sexual_orientation_table.csv",
row.names = TRUE)Check for Difference in Questionnaire Totals Between Groups
Check for Difference in IUS-27 Totals Between Groups
# t-test to check for intergroup differences in IU total score - there should
# be a significant difference.
# plot data for both groups using QQ plot
qqplot_ius_total <- ggqqplot(df, x = "ius_total",
color = "iu_group",
palette = c("#c45150", "#824372"),
title = "Q-Q Plot IUS Total") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15))
# inspect the QQ plots
qqplot_ius_total 
# save plot to file
ggsave(filename = "graphs/qqplots/qqplot_ius_total.png",
plot = qqplot_ius_total,
width = 20,
height = 10,
dpi = 300,
units = "cm")
# check significance of data for both groups using Shapiro-Wilk Test
shapiro_ius_total <- by(df$ius_total, df$iu_group, shapiro.test)
shapiro_ius_total## df$iu_group: -1
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.9603, p-value = 0.02444
##
## ------------------------------------------------------------
## df$iu_group: 1
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.88758, p-value = 0.00001673# high IU: p-value < .05, data violate assumption of normality
# low IU: p-value < .05, data violate assumption of normality
## check assumption of homogeneity of variances using Bartlett Test ##
bartlett_ius_total <- bartlett.test(ius_total ~ iu_group, data = df)
bartlett_ius_total##
## Bartlett test of homogeneity of variances
##
## data: ius_total by iu_group
## Bartlett's K-squared = 19.86, df = 1, p-value = 0.000008334# p-value < .05, data violate assumption of equal variances
## compute independent samples t.test ##
# as data violate assumption of normality and assumption of equal variances,
# use non-parametric Mann Whitney U
# compute t.test and assign values to an object
ius_total_groupdiff <- wilcox.test(ius_total ~ iu_group, data = df, paired = FALSE)
# obtain t.test values
ius_total_groupdiff##
## Wilcoxon rank sum test with continuity correction
##
## data: ius_total by iu_group
## W = 0, p-value < 0.00000000000000022
## alternative hypothesis: true location shift is not equal to 0# p-value < .05, there is a statistical difference in IUS 27 total between groupsCheck for Difference in STICSA Totals Between Groups
# t-test to check for intergroup differences in STICSA total score - there should
# be a significant difference.
# plot data for both groups using QQ plot
qqplot_sticsa_total <- ggqqplot(df, x = "sticsa_total",
color = "iu_group",
palette = c("#c45150", "#824372"),
title = "Q-Q Plot STICSA Total") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15))
# inspect the QQ plots
qqplot_sticsa_total 
# save plot to file
ggsave(filename = "graphs/qqplots/qqplot_sticsa_total.png",
plot = qqplot_sticsa_total,
width = 20,
height = 10,
dpi = 300,
units = "cm")
# check significance of data for both groups using Shapiro-Wilk Test
shapiro_sticsa_total <- by(df$sticsa_total, df$iu_group, shapiro.test)
shapiro_sticsa_total## df$iu_group: -1
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.95514, p-value = 0.01269
##
## ------------------------------------------------------------
## df$iu_group: 1
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.96452, p-value = 0.05002# high iu: p-value > .05, data meet assumption of normality
# low iu: p-value < .05, data violate assumption of normality
## check assumption of homogeneity of variances using Bartlett Test ##
bartlett_sticsa_total <- bartlett.test(sticsa_total ~ iu_group, data = df)
bartlett_sticsa_total##
## Bartlett test of homogeneity of variances
##
## data: sticsa_total by iu_group
## Bartlett's K-squared = 3.8861, df = 1, p-value = 0.04869# p-value < .05, data violate assumption of equal variances
## compute independent samples t.test ##
# as data violate assumption of normality and assumption of equal variances,
# use non-parametric Mann Whitney U
# compute t.test and assign values to an object
sticsa_total_groupdiff <- wilcox.test(sticsa_total ~ iu_group, data = df, paired = FALSE)
# obtain t.test values
sticsa_total_groupdiff##
## Wilcoxon rank sum test with continuity correction
##
## data: sticsa_total by iu_group
## W = 1032, p-value = 0.000000005691
## alternative hypothesis: true location shift is not equal to 0# p-value < .05, there is a statistical difference in STICSA total between groupsCheck for Difference in Demographics Between Groups
Check for Difference in Age Between Groups
# t-test to check for intergroup differences in age
# plot data for both groups using QQ plot
qqplot_age <- ggqqplot(df, x = "age",
color = "iu_group",
palette = c("#c45150", "#824372"),
title = "Q-Q Plot Age") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15))
# inspect the QQ plots
qqplot_age ## Warning: Removed 1 rows containing non-finite values (stat_qq).## Warning: Removed 1 rows containing non-finite values (stat_qq_line).
## Warning: Removed 1 rows containing non-finite values (stat_qq_line).
# save plot to file
ggsave(filename = "graphs/qqplots/qqplot_age.png",
plot = qqplot_age,
width = 20,
height = 10,
dpi = 300,
units = "cm")## Warning: Removed 1 rows containing non-finite values (stat_qq).
## Warning: Removed 1 rows containing non-finite values (stat_qq_line).
## Warning: Removed 1 rows containing non-finite values (stat_qq_line).# check significance of data for both groups using Shapiro-Wilk Test
shapiro_age <- by(df$age, df$iu_group, shapiro.test)
shapiro_age## df$iu_group: -1
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.95698, p-value = 0.016
##
## ------------------------------------------------------------
## df$iu_group: 1
##
## Shapiro-Wilk normality test
##
## data: dd[x, ]
## W = 0.88408, p-value = 0.00001422# high IU: p-value < .05, data violate assumption of normality
# low IU: p-value < .05, data violate assumption of normality
## check assumption of homogeneity of variances using Bartlett Test ##
bartlett_age <- bartlett.test(age ~ iu_group, data = df)
bartlett_age##
## Bartlett test of homogeneity of variances
##
## data: age by iu_group
## Bartlett's K-squared = 0.27665, df = 1, p-value = 0.5989# p-value > .05, data meet assumption of equal variances
## compute independent samples t.test ##
# as data violate assumption of normality,
# use non-parametric Mann Whitney U
# compute t.test and assign values to an object
age_groupdiff <- wilcox.test(age ~ iu_group, data = df, paired = FALSE)
# obtain t.test values
age_groupdiff##
## Wilcoxon rank sum test with continuity correction
##
## data: age by iu_group
## W = 2585.5, p-value = 0.3773
## alternative hypothesis: true location shift is not equal to 0# p-value > .05, there is no statistical difference in age between groupsCheck for Difference in Sex Between Groups
# compute chi-square of cross-tabulation and save as object
chi_sex <- chisq.test(table(df$iu_group, df$sex))
# check assumption of chi-square
chi_sex_expected <- chi_sex$expected
chi_sex_expected##
## Female Male
## -1 42.40876 27.59124
## 1 40.59124 26.40876# no cells less than 5, meets assumptions
# obtain statistic, df and p-value
chi_sex##
## Pearson's Chi-squared test with Yates' continuity correction
##
## data: table(df$iu_group, df$sex)
## X-squared = 4.2708, df = 1, p-value = 0.03877# p-value < .05, there appears to be a statistical difference in sex between groups
# therefore, obtain observed values
chi_sex_observed <- chi_sex$observed
chi_sex_observed##
## Female Male
## -1 36 34
## 1 47 20Check for Difference in Ethnicity Between Groups
# compute chi-square of cross-tabulation and save as object
chi_ethnicity <- chisq.test(table(df$iu_group, df$ethnicity)) ## Warning in chisq.test(table(df$iu_group, df$ethnicity)): Chi-squared
## approximation may be incorrect# check assumption of chi-square
chi_ethnicity$expected ##
## Asian Black Middle Eastern/ Arab Mixed White
## -1 14.5 1 2 1 43.5
## 1 14.5 1 2 1 43.5# multiple cells with values less than 5, does not meet assumptions
# and therefore requires Fisher's Exact Test
# obtain statistic and df
chi_ethnicity ##
## Pearson's Chi-squared test
##
## data: table(df$iu_group, df$ethnicity)
## X-squared = 7.7356, df = 4, p-value = 0.1018# obtain corrected p-value
chi_ethnicity_pval <- fisher.test(df$iu_group, df$ethnicity)
chi_ethnicity_pval##
## Fisher's Exact Test for Count Data
##
## data: df$iu_group and df$ethnicity
## p-value = 0.05899
## alternative hypothesis: two.sided# p-value > .05, no evidence of statistical difference in ethnicity between groupsCheck for Difference in Sexual Orientation Between Groups
# compute chi-square of cross-tabulation and save as object
chi_sexual_orientation <- chisq.test(table(df$iu_group, df$sexual_orientation))
# check assumption of chi-square
chi_sexual_orientation$expected ##
## Heterosexual Sexual Minority
## -1 51.07438 8.92562
## 1 51.92562 9.07438# no cells with values less than 5, meets assumptions
# obtain statistic and df
chi_sexual_orientation ##
## Pearson's Chi-squared test with Yates' continuity correction
##
## data: table(df$iu_group, df$sexual_orientation)
## X-squared = 0, df = 1, p-value = 1# p-value > .05, no evidence of statistical difference in sexual orientation between groupsDistribution Checks of Eye-Movement Variables
Fixation Count
Acquisition CS+
hist_acq_csp_fix_count <- df %>%
ggplot(aes(acq_csp_fix_count, fill = iu_group)) +
geom_histogram(binwidth = 1, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 30, 5)) +
labs(x = "Fixation Count", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Acquisition CS+ Fixation Count") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_acq_csp_fix_count
# save plot to file
ggsave(filename = "graphs/histograms/hist_acq_csp_fix_count.png",
plot = hist_acq_csp_fix_count,
width = 20,
height = 10,
dpi = 300,
units = "cm")Acquisition CS-
hist_acq_csm_fix_count <- df %>%
ggplot(aes(acq_csm_fix_count, fill = iu_group)) +
geom_histogram(binwidth = 1, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 30, 5)) +
labs(x = "Fixation Count", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Acquisition CS- Fixation Count") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_acq_csm_fix_count
# save plot to file
ggsave(filename = "graphs/histograms/hist_acq_csm_fix_count.png",
plot = hist_acq_csm_fix_count,
width = 20,
height = 10,
dpi = 300,
units = "cm")
# combine acquisition fixation count graphs
hists_acq_fix_count <-
grid.arrange(hist_acq_csp_fix_count, hist_acq_csm_fix_count,
ncol =2)
# save plot to file
ggsave(filename = "graphs/histograms/hists_acq_fix_count.png",
plot = hists_acq_fix_count,
width = 30,
height = 10,
dpi = 300,
units = "cm")Early Extinction CS+
hist_e_ext_csp_fix_count <- df %>%
ggplot(aes(e_ext_csp_fix_count, fill = iu_group)) +
geom_histogram(binwidth = 1, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 30, 5)) +
labs(x = "Fixation Count", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Early Extinction CS+ Fixation Count") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_e_ext_csp_fix_count
# save plot to file
ggsave(filename = "graphs/histograms/hist_e_ext_csp_fix_count.png",
plot = hist_e_ext_csp_fix_count,
width = 20,
height = 10,
dpi = 300,
units = "cm")Early Extinction CS-
hist_e_ext_csm_fix_count <- df %>%
ggplot(aes(e_ext_csm_fix_count, fill = iu_group)) +
geom_histogram(binwidth = 1, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 30, 5)) +
labs(x = "Fixation Count", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Early Extinction CS- Fixation Count") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_e_ext_csm_fix_count
# save plot to file
ggsave(filename = "graphs/histograms/hist_e_ext_csm_fix_count.png",
plot = hist_e_ext_csm_fix_count,
width = 20,
height = 10,
dpi = 300,
units = "cm")Late Extinction CS+
hist_l_ext_csp_fix_count <- df %>%
ggplot(aes(l_ext_csp_fix_count, fill = iu_group)) +
geom_histogram(binwidth = 1, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 30, 5)) +
labs(x = "Fixation Count", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Late Extinction CS+ Fixation Count") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_l_ext_csp_fix_count
# save plot to file
ggsave(filename = "graphs/histograms/hist_l_ext_csp_fix_count.png",
plot = hist_l_ext_csp_fix_count,
width = 20,
height = 10,
dpi = 300,
units = "cm")Late Extinction CS-
hist_l_ext_csm_fix_count <- df %>%
ggplot(aes(l_ext_csm_fix_count, fill = iu_group)) +
geom_histogram(binwidth = 1, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 30, 5)) +
labs(x = "Fixation Count", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Late Extinction CS- Fixation Count") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_l_ext_csm_fix_count
# save plot to file
ggsave(filename = "graphs/histograms/hist_l_ext_csm_fix_count.png",
plot = hist_l_ext_csm_fix_count,
width = 20,
height = 10,
dpi = 300,
units = "cm")
# combine extinction fixation count graphs
hists_ext_fix_count <-
grid.arrange(hist_e_ext_csp_fix_count, hist_e_ext_csm_fix_count,
hist_l_ext_csp_fix_count, hist_l_ext_csm_fix_count,
ncol =2)
# save plot to file
ggsave(filename = "graphs/histograms/hists_ext_fix_count.png",
plot = hists_ext_fix_count,
width = 30,
height = 20,
dpi = 300,
units = "cm")Fixation Duration
Acquisition CS+
hist_acq_csp_fix_duration <- df %>%
ggplot(aes(acq_csp_fix_duration, fill = iu_group)) +
geom_histogram(binwidth = 220, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 6000, 1000)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Acquisition CS+ Fixation Duration") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_acq_csp_fix_duration
# save plot to file
ggsave(filename = "graphs/histograms/hist_acq_csp_fix_duration.png",
plot = hist_acq_csp_fix_duration,
width = 20,
height = 10,
dpi = 300,
units = "cm")Acquisition CS-
hist_acq_csm_fix_duration <- df %>%
ggplot(aes(acq_csm_fix_duration, fill = iu_group)) +
geom_histogram(binwidth = 220, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 6000, 1000)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Acquisition CS- Fixation Duration") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_acq_csm_fix_duration
# save plot to file
ggsave(filename = "graphs/histograms/hist_acq_csm_fix_duration.png",
plot = hist_acq_csm_fix_duration,
width = 20,
height = 10,
dpi = 300,
units = "cm")
# combine acquisition fixation duration graphs
hists_acq_fix_duration <-
grid.arrange(hist_acq_csp_fix_duration, hist_acq_csm_fix_duration,
ncol =2)
# save plot to file
ggsave(filename = "graphs/histograms/hists_acq_fix_duration.png",
plot = hists_acq_fix_duration,
width = 30,
height = 10,
dpi = 300,
units = "cm")Early Extinction CS+
hist_e_ext_csp_fix_duration <- df %>%
ggplot(aes(e_ext_csp_fix_duration, fill = iu_group)) +
geom_histogram(binwidth = 220, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 6000, 1000)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Early Extinction CS+ Fixation Duration") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_e_ext_csp_fix_duration
# save plot to file
ggsave(filename = "graphs/histograms/hist_e_ext_csp_fix_duration.png",
plot = hist_e_ext_csp_fix_duration,
width = 20,
height = 10,
dpi = 300,
units = "cm")Early Extinction CS-
hist_e_ext_csm_fix_duration <- df %>%
ggplot(aes(e_ext_csm_fix_duration, fill = iu_group)) +
geom_histogram(binwidth = 220, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 6000, 1000)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Early Extinction CS- Fixation Duration") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_e_ext_csm_fix_duration
# save plot to file
ggsave(filename = "graphs/histograms/hist_e_ext_csm_fix_duration.png",
plot = hist_e_ext_csm_fix_duration,
width = 20,
height = 10,
dpi = 300,
units = "cm")Late Extinction CS+
hist_l_ext_csp_fix_duration <- df %>%
ggplot(aes(l_ext_csp_fix_duration, fill = iu_group)) +
geom_histogram(binwidth = 220, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 6000, 1000)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Late Extinction CS+ Fixation Duration") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_l_ext_csp_fix_duration
# save plot to file
ggsave(filename = "graphs/histograms/hist_l_ext_csp_fix_duration.png",
plot = hist_l_ext_csp_fix_duration,
width = 20,
height = 10,
dpi = 300,
units = "cm")Late Extinction CS-
hist_l_ext_csm_fix_duration <- df %>%
ggplot(aes(l_ext_csm_fix_duration, fill = iu_group)) +
geom_histogram(binwidth = 220, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 6000, 1000)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Late Extinction CS- Fixation Duration") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_l_ext_csm_fix_duration
# save plot to file
ggsave(filename = "graphs/histograms/hist_l_ext_csm_fix_duration.png",
plot = hist_l_ext_csm_fix_duration,
width = 20,
height = 10,
dpi = 300,
units = "cm")
# combine extinction fixation duration graphs
hists_ext_fix_duration <-
grid.arrange(hist_e_ext_csp_fix_duration, hist_e_ext_csm_fix_duration,
hist_l_ext_csp_fix_duration, hist_l_ext_csm_fix_duration,
ncol =2)
# save plot to file
ggsave(filename = "graphs/histograms/hists_ext_fix_duration.png",
plot = hists_ext_fix_duration,
width = 30,
height = 20,
dpi = 300,
units = "cm")Saccade Amplitude
Acquisition CS+
hist_acq_csp_sacc_amplitude <- df %>%
ggplot(aes(acq_csp_sacc_amplitude, fill = iu_group)) +
geom_histogram(binwidth = .5, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 10, 2)) +
labs(x = "Saccade Amplitude (deg/ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Acquisition CS+ Saccade Amplitude") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_acq_csp_sacc_amplitude
ggsave(filename = "graphs/histograms/hist_acq_csp_sacc_amplitude.png",
plot = hist_acq_csp_sacc_amplitude,
width = 20,
height = 10,
dpi = 300,
units = "cm")Acquisition CS-
hist_acq_csm_sacc_amplitude <- df %>%
ggplot(aes(acq_csm_sacc_amplitude, fill = iu_group)) +
geom_histogram(binwidth = .5, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 10, 2)) +
labs(x = "Saccade Amplitude (deg/ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Acquisition CS- Saccade Amplitude") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_acq_csm_sacc_amplitude## Warning: Removed 2 rows containing non-finite values (stat_bin).
ggsave(filename = "graphs/histograms/hist_acq_csm_sacc_amplitude.png",
plot = hist_acq_csm_sacc_amplitude,
width = 20,
height = 10,
dpi = 300,
units = "cm")## Warning: Removed 2 rows containing non-finite values (stat_bin).# combine acquisition saccade amplitude graphs
hists_acq_sacc_amplitude <-
grid.arrange(hist_acq_csp_sacc_amplitude, hist_acq_csm_sacc_amplitude,
ncol =2)## Warning: Removed 2 rows containing non-finite values (stat_bin).
# save plot to file
ggsave(filename = "graphs/histograms/hists_acq_sacc_amplitude.png",
plot = hists_acq_sacc_amplitude,
width = 30,
height = 10,
dpi = 300,
units = "cm")Early Extinction CS+
hist_e_ext_csp_sacc_amplitude <- df %>%
ggplot(aes(e_ext_csp_sacc_amplitude, fill = iu_group)) +
geom_histogram(binwidth = .5, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 14, 2)) +
labs(x = "Saccade Amplitude (deg/ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Early Extinction CS+ Saccade Amplitude") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_e_ext_csp_sacc_amplitude## Warning: Removed 1 rows containing non-finite values (stat_bin).
# save plot to file
ggsave(filename = "graphs/histograms/hist_e_ext_csp_sacc_amplitude.png",
plot = hist_e_ext_csp_sacc_amplitude,
width = 20,
height = 10,
dpi = 300,
units = "cm")## Warning: Removed 1 rows containing non-finite values (stat_bin).Early Extinction CS-
hist_e_ext_csm_sacc_amplitude <- df %>%
ggplot(aes(e_ext_csm_sacc_amplitude, fill = iu_group)) +
geom_histogram(binwidth = .5, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 14, 2)) +
labs(x = "Saccade Amplitude (deg/ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Early Extinction CS- Saccade Amplitude") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_e_ext_csm_sacc_amplitude## Warning: Removed 1 rows containing non-finite values (stat_bin).
# save plot to file
ggsave(filename = "graphs/histograms/hist_e_ext_csm_sacc_amplitude.png",
plot = hist_e_ext_csm_sacc_amplitude,
width = 20,
height = 10,
dpi = 300,
units = "cm")## Warning: Removed 1 rows containing non-finite values (stat_bin).Late Extinction CS+
hist_l_ext_csp_sacc_amplitude <- df %>%
ggplot(aes(l_ext_csp_sacc_amplitude, fill = iu_group)) +
geom_histogram(binwidth = .5, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 14, 2)) +
labs(x = "Saccade Amplitude (deg/ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Late Extinction CS+ Saccade Amplitude") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_l_ext_csp_sacc_amplitude## Warning: Removed 1 rows containing non-finite values (stat_bin).
# save plot to file
ggsave(filename = "graphs/histograms/hist_l_ext_csp_sacc_amplitude.png",
plot = hist_l_ext_csp_sacc_amplitude,
width = 20,
height = 10,
dpi = 300,
units = "cm")## Warning: Removed 1 rows containing non-finite values (stat_bin).Late Extinction CS-
hist_l_ext_csm_sacc_amplitude <- df %>%
ggplot(aes(l_ext_csm_sacc_amplitude, fill = iu_group)) +
geom_histogram(binwidth = .5, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 14, 2)) +
labs(x = "Saccade Amplitude (deg/ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Late Extinction CS- Saccade Amplitude") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_l_ext_csm_sacc_amplitude
# save plot to file
ggsave(filename = "graphs/histograms/hist_l_ext_csm_sacc_amplitude.png",
plot = hist_l_ext_csm_sacc_amplitude,
width = 20,
height = 10,
dpi = 300,
units = "cm")
# combine extinction saccade amplitude graphs
hists_ext_sacc_amplitude <-
grid.arrange(hist_e_ext_csp_sacc_amplitude, hist_e_ext_csm_sacc_amplitude,
hist_l_ext_csp_sacc_amplitude, hist_l_ext_csm_sacc_amplitude,
ncol =2)## Warning: Removed 1 rows containing non-finite values (stat_bin).
## Warning: Removed 1 rows containing non-finite values (stat_bin).
## Warning: Removed 1 rows containing non-finite values (stat_bin).
# save plot to file
ggsave(filename = "graphs/histograms/hists_ext_sacc_amplitude.png",
plot = hists_ext_sacc_amplitude,
width = 30,
height = 20,
dpi = 300,
units = "cm")Descriptives
Questionnaire Variables
# for all participants
descriptives_all_questionnaires <-
describe(df[, c("ius_total", "sticsa_total")], na.rm = TRUE)
# for high IU group
descriptives_high_iu_questionnaires <-
describe(df[df$iu_group == "1", c("ius_total", "sticsa_total")], na.rm = TRUE)
# for low IU group
descriptives_low_iu_questionnaires <-
describe(df[df$iu_group == "-1", c("ius_total", "sticsa_total")], na.rm = TRUE)
# combine all into table
descriptives_questionnaires_table <- round(rbind(descriptives_all_questionnaires,
descriptives_high_iu_questionnaires,
descriptives_low_iu_questionnaires), 2)
# rename rows for easier interpretation
rownames(descriptives_questionnaires_table) <- c("IUS 27 (All Participants)",
"STICSA Total (All Participants)",
"IUS 27 (High IU Group)",
"STICSA Total (High IU Group)",
"IUS 27 (Low IU Group)",
"STICSA Total (Low IU Group)")
descriptives_questionnaires_table## vars n mean sd median trimmed mad min
## IUS 27 (All Participants) 1 139 65.82 20.39 63.0 64.27 20.76 32
## STICSA Total (All Participants) 2 139 40.54 9.54 39.0 39.93 10.38 22
## IUS 27 (High IU Group) 1 68 82.65 14.77 78.0 80.79 11.86 65
## STICSA Total (High IU Group) 2 68 45.29 9.30 45.5 44.77 9.64 30
## IUS 27 (Low IU Group) 1 71 49.70 8.51 51.0 49.96 10.38 32
## STICSA Total (Low IU Group) 2 71 35.99 7.32 35.0 35.35 5.93 22
## max range skew kurtosis se
## IUS 27 (All Participants) 125 93 0.64 0.00 1.73
## STICSA Total (All Participants) 69 47 0.65 0.06 0.81
## IUS 27 (High IU Group) 125 60 1.11 0.53 1.79
## STICSA Total (High IU Group) 69 39 0.47 -0.19 1.13
## IUS 27 (Low IU Group) 64 32 -0.23 -1.05 1.01
## STICSA Total (Low IU Group) 57 35 0.76 0.30 0.87# write to csv
write.csv(descriptives_questionnaires_table, file = "tables/descriptives/descriptives_questionnaires_table.csv",
row.names = TRUE)Eye Movement Variables
Fixation Count
# for all participants
descriptives_all_fix_count <-
describe(df[, c("acq_csp_fix_count","acq_csm_fix_count",
"e_ext_csp_fix_count", "e_ext_csm_fix_count",
"l_ext_csp_fix_count", "l_ext_csm_fix_count")],
na.rm = TRUE)
# for high IU group
descriptives_high_iu_fix_count <-
describe(df[df$iu_group == "1", c("acq_csp_fix_count","acq_csm_fix_count",
"e_ext_csp_fix_count", "e_ext_csm_fix_count",
"l_ext_csp_fix_count", "l_ext_csm_fix_count")],
na.rm = TRUE)
# for low IU group
descriptives_low_iu_fix_count <-
describe(df[df$iu_group == "-1", c("acq_csp_fix_count","acq_csm_fix_count",
"e_ext_csp_fix_count", "e_ext_csm_fix_count",
"l_ext_csp_fix_count", "l_ext_csm_fix_count")],
na.rm = TRUE)
# combine all into table
descriptives_fix_count_table <- round(rbind(descriptives_all_fix_count,
descriptives_high_iu_fix_count,
descriptives_low_iu_fix_count), 2)
# rename rows for easier interpretation
rownames(descriptives_fix_count_table) <- c("Acquisition CS+ Fix Count (All Participants)",
"Acquisition CS- Fix Count (All Participants)",
"Early Extinction CS+ Fix Count (All Participants)",
"Early Extinction CS- Fix Count (All Participants)",
"Late Extinction CS+ Fix Count (All Participants)",
"Late Extinction CS- Fix Count (All Participants)",
"Acquisition CS+ Fix Count (High IU Group)",
"Acquisition CS- Fix Count (High IU Group)",
"Early Extinction CS+ Fix Count (High IU Group)",
"Early Extinction CS- Fix Count (High IU Group)",
"Late Extinction CS+ Fix Count (High IU Group)",
"Late Extinction CS- Fix Count (High IU Group)",
"Acquisition CS+ Fix Count (Low IU Group)",
"Acquisition CS- Fix Count (Low IU Group)",
"Early Extinction CS+ Fix Count (Low IU Group)",
"Early Extinction CS- Fix Count (Low IU Group)",
"Late Extinction CS+ Fix Count (Low IU Group)",
"Late Extinction CS- Fix Count (Low IU Group)")
descriptives_fix_count_table## vars n mean sd median
## Acquisition CS+ Fix Count (All Participants) 1 139 6.90 3.65 6.67
## Acquisition CS- Fix Count (All Participants) 2 139 7.31 3.25 6.75
## Early Extinction CS+ Fix Count (All Participants) 3 139 7.16 3.70 6.50
## Early Extinction CS- Fix Count (All Participants) 4 139 7.40 3.53 6.75
## Late Extinction CS+ Fix Count (All Participants) 5 139 7.55 3.49 7.25
## Late Extinction CS- Fix Count (All Participants) 6 139 7.86 3.52 7.75
## Acquisition CS+ Fix Count (High IU Group) 1 68 7.51 3.84 7.08
## Acquisition CS- Fix Count (High IU Group) 2 68 7.97 3.07 7.79
## Early Extinction CS+ Fix Count (High IU Group) 3 68 7.54 3.26 6.75
## Early Extinction CS- Fix Count (High IU Group) 4 68 8.14 3.26 7.88
## Late Extinction CS+ Fix Count (High IU Group) 5 68 8.41 3.63 7.75
## Late Extinction CS- Fix Count (High IU Group) 6 68 8.89 3.33 8.75
## Acquisition CS+ Fix Count (Low IU Group) 1 71 6.33 3.38 5.50
## Acquisition CS- Fix Count (Low IU Group) 2 71 6.67 3.31 5.92
## Early Extinction CS+ Fix Count (Low IU Group) 3 71 6.80 4.06 6.00
## Early Extinction CS- Fix Count (Low IU Group) 4 71 6.70 3.66 5.75
## Late Extinction CS+ Fix Count (Low IU Group) 5 71 6.72 3.15 6.50
## Late Extinction CS- Fix Count (Low IU Group) 6 71 6.87 3.43 6.50
## trimmed mad min max range
## Acquisition CS+ Fix Count (All Participants) 6.57 3.71 1.50 23.17 21.67
## Acquisition CS- Fix Count (All Participants) 7.06 3.71 1.92 18.33 16.42
## Early Extinction CS+ Fix Count (All Participants) 6.75 3.71 1.50 20.50 19.00
## Early Extinction CS- Fix Count (All Participants) 7.14 3.71 1.50 21.50 20.00
## Late Extinction CS+ Fix Count (All Participants) 7.33 3.34 1.00 22.00 21.00
## Late Extinction CS- Fix Count (All Participants) 7.65 3.34 1.50 20.00 18.50
## Acquisition CS+ Fix Count (High IU Group) 7.14 3.46 2.00 23.17 21.17
## Acquisition CS- Fix Count (High IU Group) 7.78 3.21 2.92 18.33 15.42
## Early Extinction CS+ Fix Count (High IU Group) 7.26 2.97 2.25 19.25 17.00
## Early Extinction CS- Fix Count (High IU Group) 8.00 3.34 2.00 16.50 14.50
## Late Extinction CS+ Fix Count (High IU Group) 8.11 3.71 1.50 22.00 20.50
## Late Extinction CS- Fix Count (High IU Group) 8.61 2.41 3.25 20.00 16.75
## Acquisition CS+ Fix Count (Low IU Group) 6.02 3.71 1.50 15.67 14.17
## Acquisition CS- Fix Count (Low IU Group) 6.35 3.21 1.92 15.50 13.58
## Early Extinction CS+ Fix Count (Low IU Group) 6.22 3.71 1.50 20.50 19.00
## Early Extinction CS- Fix Count (Low IU Group) 6.31 3.34 1.50 21.50 20.00
## Late Extinction CS+ Fix Count (Low IU Group) 6.57 3.34 1.00 16.00 15.00
## Late Extinction CS- Fix Count (Low IU Group) 6.68 3.71 1.50 17.75 16.25
## skew kurtosis se
## Acquisition CS+ Fix Count (All Participants) 1.18 2.52 0.31
## Acquisition CS- Fix Count (All Participants) 0.69 0.19 0.28
## Early Extinction CS+ Fix Count (All Participants) 1.18 1.74 0.31
## Early Extinction CS- Fix Count (All Participants) 0.82 0.96 0.30
## Late Extinction CS+ Fix Count (All Participants) 0.83 1.34 0.30
## Late Extinction CS- Fix Count (All Participants) 0.68 0.81 0.30
## Acquisition CS+ Fix Count (High IU Group) 1.43 3.48 0.47
## Acquisition CS- Fix Count (High IU Group) 0.74 0.76 0.37
## Early Extinction CS+ Fix Count (High IU Group) 1.09 1.81 0.40
## Early Extinction CS- Fix Count (High IU Group) 0.36 -0.42 0.39
## Late Extinction CS+ Fix Count (High IU Group) 1.00 1.61 0.44
## Late Extinction CS- Fix Count (High IU Group) 1.03 1.59 0.40
## Acquisition CS+ Fix Count (Low IU Group) 0.75 -0.01 0.40
## Acquisition CS- Fix Count (Low IU Group) 0.79 -0.13 0.39
## Early Extinction CS+ Fix Count (Low IU Group) 1.30 1.65 0.48
## Early Extinction CS- Fix Count (Low IU Group) 1.32 2.59 0.43
## Late Extinction CS+ Fix Count (Low IU Group) 0.47 -0.15 0.37
## Late Extinction CS- Fix Count (Low IU Group) 0.59 -0.05 0.41# write to csv
write.csv(descriptives_fix_count_table, file = "tables/descriptives/descriptives_fix_count_table.csv",
row.names = TRUE)Fixation Duration
# for all participants
descriptives_all_fix_duration <-
describe(df[, c("acq_csp_fix_duration","acq_csm_fix_duration",
"e_ext_csp_fix_duration", "e_ext_csm_fix_duration",
"l_ext_csp_fix_duration", "l_ext_csm_fix_duration")],
na.rm = TRUE)
# for high IU group
descriptives_high_iu_fix_duration <-
describe(df[df$iu_group == "1", c("acq_csp_fix_duration","acq_csm_fix_duration",
"e_ext_csp_fix_duration", "e_ext_csm_fix_duration",
"l_ext_csp_fix_duration", "l_ext_csm_fix_duration")],
na.rm = TRUE)
# for low IU group
descriptives_low_iu_fix_duration <-
describe(df[df$iu_group == "-1", c("acq_csp_fix_duration","acq_csm_fix_duration",
"e_ext_csp_fix_duration", "e_ext_csm_fix_duration",
"l_ext_csp_fix_duration", "l_ext_csm_fix_duration")],
na.rm = TRUE)
# combine all in a table
descriptives_fix_duration_table <- round(rbind(descriptives_all_fix_duration,
descriptives_high_iu_fix_duration,
descriptives_low_iu_fix_duration), 2)
# rename rows for easier interpretation
rownames(descriptives_fix_duration_table) <- c("Acquisition CS+ Fix Duration (All Participants)",
"Acquisition CS- Fix Duration (All Participants)",
"Early Extinction CS+ Fix Duration (All Participants)",
"Early Extinction CS- Fix Duration (All Participants)",
"Late Extinction CS+ Fix Duration (All Participants)",
"Late Extinction CS- Fix Duration (All Participants)",
"Acquisition CS+ Fix Duration (High IU Group)",
"Acquisition CS- Fix Duration (High IU Group)",
"Early Extinction CS+ Fix Duration (High IU Group)",
"Early Extinction CS- Fix Duration (High IU Group)",
"Late Extinction CS+ Fix Duration (High IU Group)",
"Late Extinction CS- Fix Duration (High IU Group)",
"Acquisition CS+ Fix Duration (Low IU Group)",
"Acquisition CS- Fix Duration (Low IU Group)",
"Early Extinction CS+ Fix Duration (Low IU Group)",
"Early Extinction CS- Fix Duration (Low IU Group)",
"Late Extinction CS+ Fix Duration (Low IU Group)",
"Late Extinction CS- Fix Duration (Low IU Group)")
descriptives_fix_duration_table## vars n mean sd
## Acquisition CS+ Fix Duration (All Participants) 1 139 1309.36 1173.03
## Acquisition CS- Fix Duration (All Participants) 2 139 1200.18 1048.80
## Early Extinction CS+ Fix Duration (All Participants) 3 139 1104.04 930.02
## Early Extinction CS- Fix Duration (All Participants) 4 139 1203.66 1288.87
## Late Extinction CS+ Fix Duration (All Participants) 5 139 1066.13 1094.12
## Late Extinction CS- Fix Duration (All Participants) 6 139 1068.60 1204.27
## Acquisition CS+ Fix Duration (High IU Group) 1 68 1153.24 1126.41
## Acquisition CS- Fix Duration (High IU Group) 2 68 1003.87 938.91
## Early Extinction CS+ Fix Duration (High IU Group) 3 68 869.88 621.12
## Early Extinction CS- Fix Duration (High IU Group) 4 68 833.27 912.11
## Late Extinction CS+ Fix Duration (High IU Group) 5 68 799.03 732.10
## Late Extinction CS- Fix Duration (High IU Group) 6 68 719.91 687.99
## Acquisition CS+ Fix Duration (Low IU Group) 1 71 1458.89 1204.96
## Acquisition CS- Fix Duration (Low IU Group) 2 71 1388.19 1118.70
## Early Extinction CS+ Fix Duration (Low IU Group) 3 71 1328.31 1109.79
## Early Extinction CS- Fix Duration (Low IU Group) 4 71 1558.41 1489.19
## Late Extinction CS+ Fix Duration (Low IU Group) 5 71 1321.94 1308.18
## Late Extinction CS- Fix Duration (Low IU Group) 6 71 1402.56 1474.73
## median trimmed mad
## Acquisition CS+ Fix Duration (All Participants) 789.44 1121.85 639.02
## Acquisition CS- Fix Duration (All Participants) 778.02 1017.24 606.82
## Early Extinction CS+ Fix Duration (All Participants) 786.25 958.21 611.53
## Early Extinction CS- Fix Duration (All Participants) 674.06 937.16 519.93
## Late Extinction CS+ Fix Duration (All Participants) 657.98 830.55 472.15
## Late Extinction CS- Fix Duration (All Participants) 578.00 786.22 397.29
## Acquisition CS+ Fix Duration (High IU Group) 666.40 943.03 510.01
## Acquisition CS- Fix Duration (High IU Group) 649.90 830.74 412.10
## Early Extinction CS+ Fix Duration (High IU Group) 716.65 780.80 516.83
## Early Extinction CS- Fix Duration (High IU Group) 533.31 641.57 308.46
## Late Extinction CS+ Fix Duration (High IU Group) 541.37 664.26 334.40
## Late Extinction CS- Fix Duration (High IU Group) 510.20 585.21 306.25
## Acquisition CS+ Fix Duration (Low IU Group) 1002.75 1309.89 888.32
## Acquisition CS- Fix Duration (Low IU Group) 1081.07 1216.87 977.62
## Early Extinction CS+ Fix Duration (Low IU Group) 931.86 1181.94 830.29
## Early Extinction CS- Fix Duration (Low IU Group) 1017.70 1282.48 964.06
## Late Extinction CS+ Fix Duration (Low IU Group) 781.46 1064.33 638.13
## Late Extinction CS- Fix Duration (Low IU Group) 845.97 1102.87 689.98
## min max range
## Acquisition CS+ Fix Duration (All Participants) 87.39 5083.82 4996.43
## Acquisition CS- Fix Duration (All Participants) 88.43 5446.78 5358.35
## Early Extinction CS+ Fix Duration (All Participants) 79.01 5346.50 5267.49
## Early Extinction CS- Fix Duration (All Participants) 65.23 6015.75 5950.52
## Late Extinction CS+ Fix Duration (All Participants) 121.30 5923.00 5801.70
## Late Extinction CS- Fix Duration (All Participants) 109.30 6086.56 5977.26
## Acquisition CS+ Fix Duration (High IU Group) 87.39 5083.82 4996.43
## Acquisition CS- Fix Duration (High IU Group) 88.43 5446.78 5358.35
## Early Extinction CS+ Fix Duration (High IU Group) 110.04 3044.00 2933.96
## Early Extinction CS- Fix Duration (High IU Group) 65.23 6015.75 5950.52
## Late Extinction CS+ Fix Duration (High IU Group) 121.84 4252.33 4130.49
## Late Extinction CS- Fix Duration (High IU Group) 109.30 4299.36 4190.06
## Acquisition CS+ Fix Duration (Low IU Group) 129.50 4985.33 4855.84
## Acquisition CS- Fix Duration (Low IU Group) 180.65 5219.17 5038.51
## Early Extinction CS+ Fix Duration (Low IU Group) 79.01 5346.50 5267.49
## Early Extinction CS- Fix Duration (Low IU Group) 119.97 5954.83 5834.86
## Late Extinction CS+ Fix Duration (Low IU Group) 121.30 5923.00 5801.70
## Late Extinction CS- Fix Duration (Low IU Group) 203.15 6086.56 5883.41
## skew kurtosis se
## Acquisition CS+ Fix Duration (All Participants) 1.41 1.29 99.50
## Acquisition CS- Fix Duration (All Participants) 1.65 2.73 88.96
## Early Extinction CS+ Fix Duration (All Participants) 1.58 2.76 78.88
## Early Extinction CS- Fix Duration (All Participants) 2.05 3.83 109.32
## Late Extinction CS+ Fix Duration (All Participants) 2.17 4.52 92.80
## Late Extinction CS- Fix Duration (All Participants) 2.31 4.85 102.14
## Acquisition CS+ Fix Duration (High IU Group) 1.94 3.43 136.60
## Acquisition CS- Fix Duration (High IU Group) 2.30 6.41 113.86
## Early Extinction CS+ Fix Duration (High IU Group) 1.50 2.47 75.32
## Early Extinction CS- Fix Duration (High IU Group) 3.41 14.24 110.61
## Late Extinction CS+ Fix Duration (High IU Group) 2.66 8.49 88.78
## Late Extinction CS- Fix Duration (High IU Group) 3.05 10.94 83.43
## Acquisition CS+ Fix Duration (Low IU Group) 0.98 -0.04 143.00
## Acquisition CS- Fix Duration (Low IU Group) 1.21 1.01 132.77
## Early Extinction CS+ Fix Duration (Low IU Group) 1.16 1.00 131.71
## Early Extinction CS- Fix Duration (Low IU Group) 1.41 1.11 176.73
## Late Extinction CS+ Fix Duration (Low IU Group) 1.64 1.90 155.25
## Late Extinction CS- Fix Duration (Low IU Group) 1.64 1.62 175.02# write to csv
write.csv(descriptives_fix_duration_table, file = "tables/descriptives/descriptives_fix_duration_table.csv",
row.names = TRUE)Saccade Amplitude
# for all participants
descriptives_all_sacc_amplitude <-
describe(df[, c("acq_csp_sacc_amplitude","acq_csm_sacc_amplitude",
"e_ext_csp_sacc_amplitude", "e_ext_csm_sacc_amplitude",
"l_ext_csp_sacc_amplitude", "l_ext_csm_sacc_amplitude")],
na.rm = TRUE)
# for high IU group
descriptives_high_iu_sacc_amplitude <-
describe(df[df$iu_group == "1", c("acq_csp_sacc_amplitude","acq_csm_sacc_amplitude",
"e_ext_csp_sacc_amplitude", "e_ext_csm_sacc_amplitude",
"l_ext_csp_sacc_amplitude", "l_ext_csm_sacc_amplitude")],
na.rm = TRUE)
# for low IU group
descriptives_low_iu_sacc_amplitude <-
describe(df[df$iu_group == "-1", c("acq_csp_sacc_amplitude","acq_csm_sacc_amplitude",
"e_ext_csp_sacc_amplitude", "e_ext_csm_sacc_amplitude",
"l_ext_csp_sacc_amplitude", "l_ext_csm_sacc_amplitude")],
na.rm = TRUE)
# combine all into one table
descriptives_sacc_amplitude_table <- round(rbind(descriptives_all_sacc_amplitude,
descriptives_high_iu_sacc_amplitude,
descriptives_low_iu_sacc_amplitude), 2)
# rename rows for easier interpretation
rownames(descriptives_sacc_amplitude_table) <- c("Acquisition CS+ Sacc Amplitude (All Participants)",
"Acquisition CS- Sacc Amplitude (All Participants)",
"Early Extinction CS+ Sacc Amplitude (All Participants)",
"Early Extinction CS- Sacc Amplitude (All Participants)",
"Late Extinction CS+ Sacc Amplitude (All Participants)",
"Late Extinction CS- Sacc Amplitude (All Participants)",
"Acquisition CS+ Sacc Amplitude (High IU Group)",
"Acquisition CS- Sacc Amplitude (High IU Group)",
"Early Extinction CS+ Sacc Amplitude (High IU Group)",
"Early Extinction CS- Sacc Amplitude (High IU Group)",
"Late Extinction CS+ Sacc Amplitude (High IU Group)",
"Late Extinction CS- Sacc Amplitude (High IU Group)",
"Acquisition CS+ Sacc Amplitude (Low IU Group)",
"Acquisition CS- Sacc Amplitude (Low IU Group)",
"Early Extinction CS+ Sacc Amplitude (Low IU Group)",
"Early Extinction CS- Sacc Amplitude (Low IU Group)",
"Late Extinction CS+ Sacc Amplitude (Low IU Group)",
"Late Extinction CS- Sacc Amplitude (Low IU Group)")
descriptives_sacc_amplitude_table## vars n mean sd
## Acquisition CS+ Sacc Amplitude (All Participants) 1 139 2.88 1.51
## Acquisition CS- Sacc Amplitude (All Participants) 2 137 2.98 1.57
## Early Extinction CS+ Sacc Amplitude (All Participants) 3 138 3.07 1.81
## Early Extinction CS- Sacc Amplitude (All Participants) 4 138 3.13 1.73
## Late Extinction CS+ Sacc Amplitude (All Participants) 5 138 3.00 1.92
## Late Extinction CS- Sacc Amplitude (All Participants) 6 139 3.10 1.97
## Acquisition CS+ Sacc Amplitude (High IU Group) 1 68 3.10 1.71
## Acquisition CS- Sacc Amplitude (High IU Group) 2 67 3.16 1.70
## Early Extinction CS+ Sacc Amplitude (High IU Group) 3 68 3.21 1.80
## Early Extinction CS- Sacc Amplitude (High IU Group) 4 67 3.46 1.88
## Late Extinction CS+ Sacc Amplitude (High IU Group) 5 68 3.21 1.78
## Late Extinction CS- Sacc Amplitude (High IU Group) 6 68 3.37 1.85
## Acquisition CS+ Sacc Amplitude (Low IU Group) 1 71 2.66 1.27
## Acquisition CS- Sacc Amplitude (Low IU Group) 2 70 2.80 1.43
## Early Extinction CS+ Sacc Amplitude (Low IU Group) 3 70 2.95 1.83
## Early Extinction CS- Sacc Amplitude (Low IU Group) 4 71 2.81 1.53
## Late Extinction CS+ Sacc Amplitude (Low IU Group) 5 70 2.79 2.03
## Late Extinction CS- Sacc Amplitude (Low IU Group) 6 71 2.84 2.06
## median trimmed mad min
## Acquisition CS+ Sacc Amplitude (All Participants) 2.64 2.71 1.35 0.43
## Acquisition CS- Sacc Amplitude (All Participants) 2.65 2.81 1.25 0.54
## Early Extinction CS+ Sacc Amplitude (All Participants) 2.78 2.85 1.65 0.58
## Early Extinction CS- Sacc Amplitude (All Participants) 2.66 2.94 1.35 0.42
## Late Extinction CS+ Sacc Amplitude (All Participants) 2.69 2.78 1.61 0.38
## Late Extinction CS- Sacc Amplitude (All Participants) 2.90 2.86 1.89 0.42
## Acquisition CS+ Sacc Amplitude (High IU Group) 2.99 2.92 1.74 0.43
## Acquisition CS- Sacc Amplitude (High IU Group) 2.86 2.96 1.49 0.54
## Early Extinction CS+ Sacc Amplitude (High IU Group) 3.08 3.02 1.75 0.64
## Early Extinction CS- Sacc Amplitude (High IU Group) 3.18 3.28 1.74 0.69
## Late Extinction CS+ Sacc Amplitude (High IU Group) 2.90 3.03 1.78 0.38
## Late Extinction CS- Sacc Amplitude (High IU Group) 3.13 3.22 1.90 0.61
## Acquisition CS+ Sacc Amplitude (Low IU Group) 2.52 2.56 1.17 0.59
## Acquisition CS- Sacc Amplitude (Low IU Group) 2.60 2.66 1.23 0.59
## Early Extinction CS+ Sacc Amplitude (Low IU Group) 2.48 2.70 1.52 0.58
## Early Extinction CS- Sacc Amplitude (Low IU Group) 2.34 2.65 1.07 0.42
## Late Extinction CS+ Sacc Amplitude (Low IU Group) 2.25 2.52 1.69 0.43
## Late Extinction CS- Sacc Amplitude (Low IU Group) 2.63 2.53 1.93 0.42
## max range skew
## Acquisition CS+ Sacc Amplitude (All Participants) 8.15 7.72 1.01
## Acquisition CS- Sacc Amplitude (All Participants) 8.57 8.04 1.12
## Early Extinction CS+ Sacc Amplitude (All Participants) 9.18 8.59 1.14
## Early Extinction CS- Sacc Amplitude (All Participants) 11.42 11.00 1.44
## Late Extinction CS+ Sacc Amplitude (All Participants) 13.11 12.73 1.72
## Late Extinction CS- Sacc Amplitude (All Participants) 10.95 10.53 1.25
## Acquisition CS+ Sacc Amplitude (High IU Group) 8.15 7.72 0.96
## Acquisition CS- Sacc Amplitude (High IU Group) 8.57 8.04 1.14
## Early Extinction CS+ Sacc Amplitude (High IU Group) 8.65 8.01 0.89
## Early Extinction CS- Sacc Amplitude (High IU Group) 11.42 10.73 1.38
## Late Extinction CS+ Sacc Amplitude (High IU Group) 9.62 9.24 1.11
## Late Extinction CS- Sacc Amplitude (High IU Group) 9.74 9.13 0.93
## Acquisition CS+ Sacc Amplitude (Low IU Group) 6.35 5.76 0.68
## Acquisition CS- Sacc Amplitude (Low IU Group) 7.37 6.78 0.93
## Early Extinction CS+ Sacc Amplitude (Low IU Group) 9.18 8.59 1.37
## Early Extinction CS- Sacc Amplitude (Low IU Group) 9.11 8.69 1.33
## Late Extinction CS+ Sacc Amplitude (Low IU Group) 13.11 12.68 2.20
## Late Extinction CS- Sacc Amplitude (Low IU Group) 10.95 10.53 1.57
## kurtosis se
## Acquisition CS+ Sacc Amplitude (All Participants) 0.86 0.13
## Acquisition CS- Sacc Amplitude (All Participants) 1.35 0.13
## Early Extinction CS+ Sacc Amplitude (All Participants) 1.13 0.15
## Early Extinction CS- Sacc Amplitude (All Participants) 3.30 0.15
## Late Extinction CS+ Sacc Amplitude (All Participants) 5.31 0.16
## Late Extinction CS- Sacc Amplitude (All Participants) 1.95 0.17
## Acquisition CS+ Sacc Amplitude (High IU Group) 0.40 0.21
## Acquisition CS- Sacc Amplitude (High IU Group) 1.13 0.21
## Early Extinction CS+ Sacc Amplitude (High IU Group) 0.32 0.22
## Early Extinction CS- Sacc Amplitude (High IU Group) 3.07 0.23
## Late Extinction CS+ Sacc Amplitude (High IU Group) 1.55 0.22
## Late Extinction CS- Sacc Amplitude (High IU Group) 0.76 0.22
## Acquisition CS+ Sacc Amplitude (Low IU Group) -0.16 0.15
## Acquisition CS- Sacc Amplitude (Low IU Group) 0.81 0.17
## Early Extinction CS+ Sacc Amplitude (Low IU Group) 1.92 0.22
## Early Extinction CS- Sacc Amplitude (Low IU Group) 2.40 0.18
## Late Extinction CS+ Sacc Amplitude (Low IU Group) 7.87 0.24
## Late Extinction CS- Sacc Amplitude (Low IU Group) 3.05 0.25# write to csv
write.csv(descriptives_sacc_amplitude_table, file = "tables/descriptives/descriptives_sacc_amplitude_table.csv",
row.names = TRUE)Data Transformation
Log-Transformation of Fixation Duration
# as fixation duration had high skew (>3) in high IU group for early and late
# extinction CS-, fixation duration will be log-transformed for each condition
# for acquisition CS+
df$acq_csp_fix_duration_log <- log(df$acq_csp_fix_duration)
# for acquisition CS-
df$acq_csm_fix_duration_log <- log(df$acq_csm_fix_duration)
# for early extinction CS+
df$e_ext_csp_fix_duration_log <- log(df$e_ext_csp_fix_duration)
# for early extinction CS-
df$e_ext_csm_fix_duration_log <- log(df$e_ext_csm_fix_duration)
# for late extinction CS+
df$l_ext_csp_fix_duration_log <- log(df$l_ext_csp_fix_duration)
# for late extinction CS-
df$l_ext_csm_fix_duration_log <- log(df$l_ext_csm_fix_duration)Check Descriptives of Fixation Duration Following Log-Transformation
# re-compute descriptives for fixation duration following log transformation
# for all participants
descriptives_all_fix_duration_log <-
describe(df[, c("acq_csp_fix_duration_log","acq_csm_fix_duration_log",
"e_ext_csp_fix_duration_log", "e_ext_csm_fix_duration_log",
"l_ext_csp_fix_duration_log", "l_ext_csm_fix_duration_log")],
na.rm = TRUE)
# for high IU group
descriptives_high_iu_fix_duration_log <-
describe(df[df$iu_group == "1", c("acq_csp_fix_duration_log","acq_csm_fix_duration_log",
"e_ext_csp_fix_duration_log", "e_ext_csm_fix_duration_log",
"l_ext_csp_fix_duration_log", "l_ext_csm_fix_duration_log")],
na.rm = TRUE)
# for low IU group
descriptives_low_iu_fix_duration_log <-
describe(df[df$iu_group == "-1", c("acq_csp_fix_duration_log","acq_csm_fix_duration_log",
"e_ext_csp_fix_duration_log", "e_ext_csm_fix_duration_log",
"l_ext_csp_fix_duration_log", "l_ext_csm_fix_duration_log")],
na.rm = TRUE)
# combine all to table
descriptives_fix_duration_table_log <- round(rbind(descriptives_all_fix_duration_log,
descriptives_high_iu_fix_duration_log,
descriptives_low_iu_fix_duration_log), 2)
# rename rows for easier interpretation
rownames(descriptives_fix_duration_table_log) <- c("Acquisition CS+ Fix Duration (All Participants)",
"Acquisition CS- Fix Duration (All Participants)",
"Early Extinction CS+ Fix Duration (All Participants)",
"Early Extinction CS- Fix Duration (All Participants)",
"Late Extinction CS+ Fix Duration (All Participants)",
"Late Extinction CS- Fix Duration (All Participants)",
"Acquisition CS+ Fix Duration (High IU Group)",
"Acquisition CS- Fix Duration (High IU Group)",
"Early Extinction CS+ Fix Duration (High IU Group)",
"Early Extinction CS- Fix Duration (High IU Group)",
"Late Extinction CS+ Fix Duration (High IU Group)",
"Late Extinction CS- Fix Duration (High IU Group)",
"Acquisition CS+ Fix Duration (Low IU Group)",
"Acquisition CS- Fix Duration (Low IU Group)",
"Early Extinction CS+ Fix Duration (Low IU Group)",
"Early Extinction CS- Fix Duration (Low IU Group)",
"Late Extinction CS+ Fix Duration (Low IU Group)",
"Late Extinction CS- Fix Duration (Low IU Group)")
descriptives_fix_duration_table_log## vars n mean sd median
## Acquisition CS+ Fix Duration (All Participants) 1 139 6.80 0.89 6.67
## Acquisition CS- Fix Duration (All Participants) 2 139 6.76 0.83 6.66
## Early Extinction CS+ Fix Duration (All Participants) 3 139 6.68 0.84 6.67
## Early Extinction CS- Fix Duration (All Participants) 4 139 6.67 0.90 6.51
## Late Extinction CS+ Fix Duration (All Participants) 5 139 6.60 0.83 6.49
## Late Extinction CS- Fix Duration (All Participants) 6 139 6.56 0.85 6.36
## Acquisition CS+ Fix Duration (High IU Group) 1 68 6.68 0.87 6.50
## Acquisition CS- Fix Duration (High IU Group) 2 68 6.60 0.78 6.48
## Early Extinction CS+ Fix Duration (High IU Group) 3 68 6.54 0.70 6.57
## Early Extinction CS- Fix Duration (High IU Group) 4 68 6.40 0.75 6.28
## Late Extinction CS+ Fix Duration (High IU Group) 5 68 6.41 0.72 6.29
## Late Extinction CS- Fix Duration (High IU Group) 6 68 6.31 0.71 6.23
## Acquisition CS+ Fix Duration (Low IU Group) 1 71 6.92 0.89 6.91
## Acquisition CS- Fix Duration (Low IU Group) 2 71 6.91 0.85 6.99
## Early Extinction CS+ Fix Duration (Low IU Group) 3 71 6.81 0.94 6.84
## Early Extinction CS- Fix Duration (Low IU Group) 4 71 6.92 0.96 6.93
## Late Extinction CS+ Fix Duration (Low IU Group) 5 71 6.79 0.89 6.66
## Late Extinction CS- Fix Duration (Low IU Group) 6 71 6.81 0.91 6.74
## trimmed mad min max
## Acquisition CS+ Fix Duration (All Participants) 6.80 0.94 4.47 8.53
## Acquisition CS- Fix Duration (All Participants) 6.74 0.97 4.48 8.60
## Early Extinction CS+ Fix Duration (All Participants) 6.69 0.93 4.37 8.58
## Early Extinction CS- Fix Duration (All Participants) 6.63 0.84 4.18 8.70
## Late Extinction CS+ Fix Duration (All Participants) 6.55 0.86 4.80 8.69
## Late Extinction CS- Fix Duration (All Participants) 6.50 0.71 4.69 8.71
## Acquisition CS+ Fix Duration (High IU Group) 6.67 0.79 4.47 8.53
## Acquisition CS- Fix Duration (High IU Group) 6.57 0.74 4.48 8.60
## Early Extinction CS+ Fix Duration (High IU Group) 6.54 0.77 4.70 8.02
## Early Extinction CS- Fix Duration (High IU Group) 6.35 0.67 4.18 8.70
## Late Extinction CS+ Fix Duration (High IU Group) 6.38 0.75 4.80 8.36
## Late Extinction CS- Fix Duration (High IU Group) 6.29 0.61 4.69 8.37
## Acquisition CS+ Fix Duration (Low IU Group) 6.94 1.15 4.86 8.51
## Acquisition CS- Fix Duration (Low IU Group) 6.91 1.01 5.20 8.56
## Early Extinction CS+ Fix Duration (Low IU Group) 6.86 1.15 4.37 8.58
## Early Extinction CS- Fix Duration (Low IU Group) 6.91 1.14 4.79 8.69
## Late Extinction CS+ Fix Duration (Low IU Group) 6.75 0.91 4.80 8.69
## Late Extinction CS- Fix Duration (Low IU Group) 6.75 0.96 5.31 8.71
## range skew kurtosis se
## Acquisition CS+ Fix Duration (All Participants) 4.06 0.02 -0.64 0.08
## Acquisition CS- Fix Duration (All Participants) 4.12 0.11 -0.65 0.07
## Early Extinction CS+ Fix Duration (All Participants) 4.21 -0.11 -0.40 0.07
## Early Extinction CS- Fix Duration (All Participants) 4.52 0.34 -0.31 0.08
## Late Extinction CS+ Fix Duration (All Participants) 3.89 0.45 -0.27 0.07
## Late Extinction CS- Fix Duration (All Participants) 4.02 0.59 -0.02 0.07
## Acquisition CS+ Fix Duration (High IU Group) 4.06 0.09 -0.21 0.11
## Acquisition CS- Fix Duration (High IU Group) 4.12 0.25 -0.01 0.09
## Early Extinction CS+ Fix Duration (High IU Group) 3.32 -0.08 -0.48 0.08
## Early Extinction CS- Fix Duration (High IU Group) 4.52 0.48 1.07 0.09
## Late Extinction CS+ Fix Duration (High IU Group) 3.55 0.41 -0.03 0.09
## Late Extinction CS- Fix Duration (High IU Group) 3.67 0.35 0.66 0.09
## Acquisition CS+ Fix Duration (Low IU Group) 3.65 -0.07 -1.03 0.11
## Acquisition CS- Fix Duration (Low IU Group) 3.36 -0.07 -1.05 0.10
## Early Extinction CS+ Fix Duration (Low IU Group) 4.21 -0.32 -0.53 0.11
## Early Extinction CS- Fix Duration (Low IU Group) 3.90 0.03 -0.87 0.11
## Late Extinction CS+ Fix Duration (Low IU Group) 3.89 0.29 -0.67 0.11
## Late Extinction CS- Fix Duration (Low IU Group) 3.40 0.47 -0.83 0.11# write to csv
write.csv(descriptives_fix_duration_table_log, file = "tables/descriptives/descriptives_fix_duration_table_log.csv",
row.names = TRUE)
### there are no longer any skew values of +/- 3. Check Histograms of Fixation Duration Following Log-Transformation
Acquisition CS+
########## pre-log-transformation
hist_acq_csp_fix_duration
########## post-log-transformation
hist_acq_csp_fix_duration_log <- df %>%
ggplot(aes(acq_csp_fix_duration_log, fill = iu_group)) +
geom_histogram(binwidth = .2, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 12, 2)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Acquisition CS+ Fixation Duration (Log-Transformed)") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_acq_csp_fix_duration_log
# save plot to file
ggsave(filename = "graphs/histograms/hist_acq_csp_fix_duration_log.png",
plot = hist_acq_csp_fix_duration_log,
width = 20,
height = 10,
dpi = 300,
units = "cm")Acquisition CS-
########## pre-log-transformation
hist_acq_csp_fix_duration_log
########## post-log-transformation
hist_acq_csm_fix_duration_log <- df %>%
ggplot(aes(acq_csm_fix_duration_log, fill = iu_group)) +
geom_histogram(binwidth = .2, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 12, 2)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Acquisition CS- Fixation Duration (Log-Transformed)") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_acq_csm_fix_duration_log
# save plot to file
ggsave(filename = "graphs/histograms/hist_acq_csm_fix_duration_log.png",
plot = hist_acq_csm_fix_duration_log,
width = 20,
height = 10,
dpi = 300,
units = "cm")
# combine histograms of acquisition fix duration pre and post log-transformation
hists_fix_duration_acq_log <- grid.arrange(hist_acq_csp_fix_duration, hist_acq_csm_fix_duration,
hist_acq_csp_fix_duration_log, hist_acq_csm_fix_duration_log,
ncol = 2)
# save plot to file
ggsave(filename = "graphs/histograms/hists_fix_duration_acq_log.png",
plot = hists_fix_duration_acq_log,
width = 30,
height = 20,
dpi = 300,
units = "cm")Early Extinction CS+
########## pre-log-transformation
hist_e_ext_csp_fix_duration
########## post-log-transformation
hist_e_ext_csp_fix_duration_log <- df %>%
ggplot(aes(e_ext_csp_fix_duration_log, fill = iu_group)) +
geom_histogram(binwidth = .2, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 12, 2)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Early Extinction CS+ Fixation Duration (Log-Transformed)") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_e_ext_csp_fix_duration_log
# save plot to file
ggsave(filename = "graphs/histograms/hist_e_ext_csp_fix_duration_log.png",
plot = hist_e_ext_csp_fix_duration_log,
width = 20,
height = 10,
dpi = 300,
units = "cm")Early Extinction CS-
########## pre-log-transformation
hist_e_ext_csm_fix_duration
########## post-log-transformation
hist_e_ext_csm_fix_duration_log <- df %>%
ggplot(aes(e_ext_csm_fix_duration_log, fill = iu_group)) +
geom_histogram(binwidth = .2, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 12, 2)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Early Extinction CS- Fixation Duration (Log-Transformed)") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_e_ext_csm_fix_duration_log
# save plot to file
ggsave(filename = "graphs/histograms/hist_e_ext_csm_fix_duration_log.png",
plot = hist_e_ext_csm_fix_duration_log,
width = 20,
height = 10,
dpi = 300,
units = "cm")
# combine histograms of early extinction fix duration pre and post log-transformation
hists_fix_duration_e_ext_log <- grid.arrange(hist_e_ext_csp_fix_duration, hist_e_ext_csm_fix_duration,
hist_e_ext_csp_fix_duration_log, hist_e_ext_csm_fix_duration_log,
ncol = 2)
# save plot to file
ggsave(filename = "graphs/histograms/hists_fix_duration_e_ext_log.png",
plot = hists_fix_duration_e_ext_log,
width = 30,
height = 20,
dpi = 300,
units = "cm")Late Extinction CS+
########## pre-log-transformation
hist_l_ext_csp_fix_duration
########## post-log-transformation
hist_l_ext_csp_fix_duration_log <- df %>%
ggplot(aes(l_ext_csp_fix_duration_log, fill = iu_group)) +
geom_histogram(binwidth = .2, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 12, 2)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Late Extinction CS+ Fixation Duration (Log-Transformed)") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_l_ext_csp_fix_duration_log
# save plot to file
ggsave(filename = "graphs/histograms/hist_l_ext_csp_fix_duration_log.png",
plot = hist_l_ext_csp_fix_duration_log,
width = 20,
height = 10,
dpi = 300,
units = "cm")Late Extinction CS-
########## pre-log-transformation
hist_l_ext_csm_fix_duration
########## post-log-transformation
hist_l_ext_csm_fix_duration_log <- df %>%
ggplot(aes(l_ext_csm_fix_duration_log, fill = iu_group)) +
geom_histogram(binwidth = .2, colour = "white", alpha = .5, position = "identity") +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
scale_x_continuous(breaks = seq(0, 12, 2)) +
labs(x = "Fixation Duration (ms)", y = "Frequency") +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
ggtitle("Histogram Late Extinction CS- Fixation Duration (Log-Transformed)") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
guides(fill = guide_legend(reverse = TRUE)) +
scale_fill_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(fill = "IU Group")
hist_l_ext_csm_fix_duration_log
# save plot to file
ggsave(filename = "graphs/histograms/hist_l_ext_csm_fix_duration_log.png",
plot = hist_l_ext_csm_fix_duration_log,
width = 20,
height = 10,
dpi = 300,
units = "cm")
# combine histograms of late extinction fix duration pre and post log-transformation
hists_fix_duration_l_ext_log <- grid.arrange(hist_l_ext_csp_fix_duration, hist_l_ext_csm_fix_duration,
hist_l_ext_csp_fix_duration_log, hist_l_ext_csm_fix_duration_log,
ncol = 2)
# save plot to file
ggsave(filename = "graphs/histograms/hists_fix_duration_l_ext_log.png",
plot = hists_fix_duration_l_ext_log,
width = 30,
height = 20,
dpi = 300,
units = "cm")ANOVAs
ANOVA Acquisition Fixation Count
# transform wide format data into long format for mixed ANOVA
df_long_acq_fix_count <- melt(df, id = c("id", "iu_group"),
measure.vars = c("acq_csp_fix_count",
"acq_csm_fix_count"))
# rename columns for easier interpretation
colnames(df_long_acq_fix_count) = c("id", "iu_group", "condition", "fix_count")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_acq_fix_count$stimulus <-
factor(ifelse(df_long_acq_fix_count$condition == "acq_csp_fix_count", 1, -1))
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) mixed ANOVA,
# and obtain effect size (partial eta squared)
acq_fix_count_anova <-
anova_test(df_long_acq_fix_count, fix_count ~ iu_group * stimulus + Error(id/stimulus),
effect.size = "pes")
# obtain the mixed ANOVA results
get_anova_table(acq_fix_count_anova)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 iu_group 1 137 4.806 0.030000 * 0.034
## 2 stimulus 1 137 11.441 0.000937 * 0.077
## 3 iu_group:stimulus 1 137 0.258 0.613000 0.002# results:
# IU: F(1,137) = 4.81, p = .030*, eta2(partial) = .034
# Stimulus: F(1,137) = 11.44, p < .001***, eta2(partial) = .077
# IU * Stimulus: F(1, 137) = 0.26, p = .613, eta2(partial) = .002
# therefore, there is a significant effect of IU & Stimulus on fixation count in acquisition,
# and no significant IU*Stimulus interaction
# write to csv
write.csv((get_anova_table(acq_fix_count_anova)),
file = "tables/anovas/acq_fix_count_anova.csv")ANOVA Acquisition Fixation Duration (Log Transformed)
# transform wide format data into long format for mixed ANOVA
df_long_acq_fix_duration_log <- melt(df, id = c("id", "iu_group"),
measure.vars = c("acq_csp_fix_duration_log",
"acq_csm_fix_duration_log"))
# the first input for melt command is the df with wide format. Second input
# is id =, which is where we list ppts with specific variables within wide format
# df. Here we have ppts ID no's as participant specific variable, and IU
# group they are assigned to is also specific for each participant. Therefore
# only need to list the two variables after id =.
# rename columns for easier interpretation
colnames(df_long_acq_fix_duration_log) = c("id", "iu_group", "condition", "fix_duration_log")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_acq_fix_duration_log$stimulus <-
factor(ifelse(df_long_acq_fix_duration_log$condition == "acq_csp_fix_duration_log", 1, -1))
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) mixed ANOVA,
# and obtain effect size (partial eta squared)
acq_fix_duration_anova_log <-
anova_test(df_long_acq_fix_duration_log, fix_duration_log ~ iu_group * stimulus + Error(id/stimulus),
effect.size = "pes")
# the error(id/stimulus) variable is unique to repeated-measures ANOVA, and means
# that the variable 'stimulus' is manipulated within 'id'
# obtain the mixed ANOVA results
get_anova_table(acq_fix_duration_anova_log)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 iu_group 1 137 3.907 0.050 0.028
## 2 stimulus 1 137 2.921 0.090 0.021
## 3 iu_group:stimulus 1 137 1.271 0.261 0.009# results:
# IU: F(1,137) = 3.91, p = .050*, eta2(partial) = .028
# Stimulus: F(1,137) = 2.92, p = .090, eta2(partial) = .021
# IU * Stimulus: F(1, 137) = 1.27, p = .261, eta2(partial) = .009
# therefore, there is a sig effect of IU, and no
# sig effect of stimulus or IU-stimulus interaction
# write to csv
write.csv((get_anova_table(acq_fix_duration_anova_log)),
file = "tables/anovas/acq_fix_duration_anova_log.csv")ANOVA Acquisition Saccade Amplitude
# transform wide format data into long format for mixed ANOVA
df_long_acq_sacc_amplitude <- melt(df, id = c("id", "iu_group"),
measure.vars = c("acq_csp_sacc_amplitude",
"acq_csm_sacc_amplitude"))
# rename columns for easier interpretation
colnames(df_long_acq_sacc_amplitude) = c("id", "iu_group", "condition", "sacc_amplitude")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_acq_sacc_amplitude$stimulus <-
factor(ifelse(df_long_acq_sacc_amplitude$condition == "acq_csp_sacc_amplitude", 1, -1))
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) mixed ANOVA,
# and obtain effect size (partial eta squared)
acq_sacc_amplitude_anova <-
anova_test(df_long_acq_sacc_amplitude, sacc_amplitude ~ iu_group * stimulus + Error(id/stimulus),
effect.size = "pes") ## Warning: NA detected in rows: 234,259.
## Removing this rows before the analysis.# obtain the mixed ANOVA results
get_anova_table(acq_sacc_amplitude_anova)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 iu_group 1 135 2.984 0.086 0.022
## 2 stimulus 1 135 0.950 0.332 0.007
## 3 iu_group:stimulus 1 135 0.379 0.539 0.003# results:
# IU: F(1,135) = 2.98, p = .086, eta2(partial) = .022
# Stimulus: F(1,135) = 0.95, p = .332, eta2(partial) = .007
# IU * Stimulus: F(1, 135) = 0.38, p = .539, eta2(partial) = .003
# therefore, there are no significant effects on saccade amplitude in
# acquisition
# write to csv
write.csv((get_anova_table(acq_sacc_amplitude_anova)),
file = "tables/anovas/acq_sacc_amplitude_anova.csv")ANOVA Extinction Fixation Count
# transform wide format data into long format for mixed ANOVA
df_long_ext_fix_count <- melt(df, id = c("id", "iu_group"),
measure.vars = c("e_ext_csp_fix_count",
"e_ext_csm_fix_count",
"l_ext_csp_fix_count",
"l_ext_csm_fix_count"))
# rename columns for easier interpretation
colnames(df_long_ext_fix_count) = c("id", "iu_group", "condition", "fix_count")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_ext_fix_count$stimulus <-
factor(ifelse(df_long_ext_fix_count$condition == "e_ext_csp_fix_count" |
df_long_ext_fix_count$condition == "l_ext_csp_fix_count", 1, -1))
# create column to code extinction as early (1) and late (-1)
df_long_ext_fix_count$time <-
factor(ifelse(df_long_ext_fix_count$condition == "e_ext_csp_fix_count" |
df_long_ext_fix_count$condition == "e_ext_csm_fix_count", 1, -1))
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) x 2 (Time: Early, Late) mixed ANOVA,
# and obtain effect size (partial eta squared)
ext_fix_count_anova <-
anova_test(df_long_ext_fix_count,
fix_count ~ iu_group * stimulus * time + Error(id/(stimulus*time)),
effect.size = "pes")
# obtain the mixed ANOVA results
get_anova_table(ext_fix_count_anova)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 iu_group 1 137 7.672 0.006 * 0.053000
## 2 stimulus 1 137 4.155 0.043 * 0.029000
## 3 time 1 137 5.733 0.018 * 0.040000
## 4 iu_group:stimulus 1 137 3.460 0.065 0.025000
## 5 iu_group:time 1 137 4.572 0.034 * 0.032000
## 6 stimulus:time 1 137 0.061 0.806 0.000443
## 7 iu_group:stimulus:time 1 137 0.600 0.440 0.004000# results:
# IU: F(1,137) = 7.67, p = .006 ***, eta2(partial) = .053
# Stimulus: F(1,137) = 4.16, p = .043 *, eta2(partial) = .029
# Time: F(1,137) = 5.73, p = .018 *, eta2(partial) = .049
# IU * Stimulus: F(1, 137) = 3.46, p = .065, eta2(partial) = .025
# IU * Time: F(1,137) = 4.57, p = .034 *, eta2(partial) = . 032
# Stimulus * Time: F(1,137) = 0.06, p = .806, eta2(partial) < .001
# IU * Stimulus * Time: F(1,137) = 0.60, p = .440, eta2(partial) = .004
# therefore, there is a significant effect of IU, Stimulus and Time on fixation count in extinction,
# as well as a significant interaction effect of IU * Time,
# but no other significant interactions.
# write to csv
write.csv((get_anova_table(ext_fix_count_anova)),
file = "tables/anovas/ext_fix_count_anova.csv")
# as there was a significant IU*Time interaction, conduct simple
# main effects analysis:
## obtain effect of IU at each level of time
simple_effects_ext_fix_count_iu <- df_long_ext_fix_count %>%
group_by(time) %>%
anova_test(dv = fix_count, wid = id, between = iu_group, within = stimulus, effect.size = "pes") %>%
get_anova_table()
# get the output
simple_effects_ext_fix_count_iu## # A tibble: 6 × 8
## time Effect DFn DFd F p `p<.05` pes
## * <fct> <chr> <dbl> <dbl> <dbl> <dbl> <chr> <dbl>
## 1 -1 iu_group 1 137 11.4 0.000952 "*" 0.077
## 2 -1 stimulus 1 137 3.38 0.068 "" 0.024
## 3 -1 iu_group:stimulus 1 137 0.864 0.354 "" 0.006
## 4 1 iu_group 1 137 3.63 0.059 "" 0.026
## 5 1 stimulus 1 137 1.50 0.222 "" 0.011
## 6 1 iu_group:stimulus 1 137 3.04 0.084 "" 0.022# results:
# The effect of IU group at early extinction was not significant [F(1,137) = 3.63, p = .059, pes = .026]
# the effect of IU group at late extinction was significant [F(1,137) = 11.41, p < .001, pes = .077]ANOVA Extinction Fixation Duration (Log Transformed)
# transform wide format data into long format for mixed ANOVA
df_long_ext_fix_duration_log <- melt(df, id = c("id", "iu_group"),
measure.vars = c("e_ext_csp_fix_duration_log",
"e_ext_csm_fix_duration_log",
"l_ext_csp_fix_duration_log",
"l_ext_csm_fix_duration_log"))
# rename columns for easier interpretation
colnames(df_long_ext_fix_duration_log) = c("id", "iu_group", "condition", "fix_duration_log")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_ext_fix_duration_log$stimulus <-
factor(ifelse(df_long_ext_fix_duration_log$condition == "e_ext_csp_fix_duration_log" |
df_long_ext_fix_duration_log$condition == "l_ext_csp_fix_duration_log", 1, -1))
# create column to code extinction as early (1) and late (-1)
df_long_ext_fix_duration_log$time <-
factor(ifelse(df_long_ext_fix_duration_log$condition == "e_ext_csp_fix_duration_log" |
df_long_ext_fix_duration_log$condition == "e_ext_csm_fix_duration_log", 1, -1))
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) x 2 (Time: Early, Late) mixed ANOVA,
# and obtain effect size (partial eta squared)
ext_fix_duration_anova_log <-
anova_test(df_long_ext_fix_duration_log,
fix_duration_log ~ iu_group * stimulus * time + Error(id/(stimulus*time)),
effect.size = "pes")
# obtain the mixed ANOVA results
get_anova_table(ext_fix_duration_anova_log)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 iu_group 1 137 11.213 0.001 * 0.076
## 2 stimulus 1 137 0.510 0.477 0.004
## 3 time 1 137 4.351 0.039 * 0.031
## 4 iu_group:stimulus 1 137 5.823 0.017 * 0.041
## 5 iu_group:time 1 137 0.241 0.624 0.002
## 6 stimulus:time 1 137 0.171 0.680 0.001
## 7 iu_group:stimulus:time 1 137 0.946 0.333 0.007# results:
# IU: F(1,137) = 11.21, p < .001 *, eta2(partial) = .076
# Stimulus: F(1,137) = 0.51, p = .477, eta2(partial) = .004
# Time: F(1,137) = 4.35, p = .039*, eta2(partial) = .031
# IU * Stimulus: F(1, 137) = 5.82, p = .017*, eta2(partial) = .041
# IU * Time: F(1,137) = 0.24, p = .624, eta2(partial) = .002
# Stimulus * Time: F(1,137) = 0.17, p = 680, eta2(partial) = .001
# IU * Stimulus * Time: F(1,137) = 0.95, p = .333, eta2(partial) = .007
# therefore, there is a significant effect of IU, Time and IU-Stimulus
# interaction on fixation duration in extinction,
# and no other significant effects or interactions.
# write to csv
write.csv((get_anova_table(ext_fix_duration_anova_log)),
file = "tables/anovas/ext_fix_duration_anova_log.csv")
# as there was a significant IU*Stimulus interaction, conduct simple
# main effects analysis:
## obtain effect of IU at each level of stimulus
simple_effects_ext_fix_duration_log_iu <- df_long_ext_fix_duration_log %>%
group_by(stimulus) %>%
anova_test(dv = fix_duration_log, wid = id, between = iu_group, within = time, effect.size = "pes") %>%
get_anova_table()
# get the output
simple_effects_ext_fix_duration_log_iu## # A tibble: 6 × 8
## stimulus Effect DFn DFd F p `p<.05` pes
## * <fct> <chr> <dbl> <dbl> <dbl> <dbl> <chr> <dbl>
## 1 -1 iu_group 1 137 14.4 0.000218 "*" 0.095
## 2 -1 time 1 137 4.34 0.039 "*" 0.031
## 3 -1 iu_group:time 1 137 0.026 0.871 "" 0.000192
## 4 1 iu_group 1 137 6.70 0.011 "*" 0.047
## 5 1 time 1 137 1.94 0.166 "" 0.014
## 6 1 iu_group:time 1 137 0.816 0.368 "" 0.006# results:
# The effect of IU group in response to CS+ was significant [F(1,137) = 6.70, p = .011, pes = .047]
# the effect of IU group in response to CS- was also significant [F(1,137) = 14.43, p < .001, pes = .095]ANOVA Extinction Saccade Amplitude
# transform wide format data into long format for mixed ANOVA
df_long_ext_sacc_amplitude <- melt(df, id = c("id", "iu_group"),
measure.vars = c("e_ext_csp_sacc_amplitude",
"e_ext_csm_sacc_amplitude",
"l_ext_csp_sacc_amplitude",
"l_ext_csm_sacc_amplitude"))
# rename columns for easier interpretation
colnames(df_long_ext_sacc_amplitude) = c("id", "iu_group", "condition", "sacc_amplitude")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_ext_sacc_amplitude$stimulus <-
factor(ifelse(df_long_ext_sacc_amplitude$condition == "e_ext_csp_sacc_amplitude" |
df_long_ext_sacc_amplitude$condition == "l_ext_csp_sacc_amplitude", 1, -1))
# create column to code extinction as early (1) and late (-1)
df_long_ext_sacc_amplitude$time <-
factor(ifelse(df_long_ext_sacc_amplitude$condition == "e_ext_csp_sacc_amplitude" |
df_long_ext_sacc_amplitude$condition == "e_ext_csm_sacc_amplitude", 1, -1))
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) x 2 (Time: Early, Late) mixed ANOVA,
# and obtain effect size (partial eta squared)
ext_sacc_amplitude_anova <-
anova_test(df_long_ext_sacc_amplitude,
sacc_amplitude ~ iu_group * stimulus * time + Error(id/(stimulus*time)),
effect.size = "pes") ## Warning: NA detected in rows: 116,181,301.
## Removing this rows before the analysis.# obtain the mixed ANOVA results
get_anova_table(ext_sacc_amplitude_anova)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 iu_group 1 134 3.170 0.077 0.023000
## 2 stimulus 1 134 0.740 0.391 0.005000
## 3 time 1 134 0.275 0.601 0.002000
## 4 iu_group:stimulus 1 134 1.687 0.196 0.012000
## 5 iu_group:time 1 134 0.131 0.718 0.000977
## 6 stimulus:time 1 134 0.077 0.781 0.000577
## 7 iu_group:stimulus:time 1 134 0.609 0.437 0.005000# results:
# IU: F(1,134) = 3.17, p = .077, eta2(partial) = .023
# Stimulus: F(1,134) = 0.74, p = .391, eta2(partial) = .005
# Time: F(1,134) = 0.28, p = .601, eta2(partial) = .002
# IU * Stimulus: F(1, 134) = 1.69, p = .196, eta2(partial) = .012
# IU * Time: F(1,134) = 0.13, p = .718, eta2(partial) < .001
# Stimulus * Time: F(1,134) = 0.08, p = .781, eta2(partial) < .001
# IU * Stimulus * Time: F(1,134) = .61, p = .437, eta2(partial) < .001
# therefore, there are no significant effects or interactions
# on saccade amplitude throughout extinction
# write to csv
write.csv((get_anova_table(ext_sacc_amplitude_anova)),
file = "tables/anovas/ext_sacc_amplitude_anova.csv")Bar Graphs
Fixation Count
Acquisition
# obtain mean fix count for each group at each stimulus type and save as vector
mean_acq_fix_count_high_iu_csp <-
mean(df$acq_csp_fix_count[df_long_acq_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS+
mean_acq_fix_count_low_iu_csp <-
mean(df$acq_csp_fix_count[df_long_acq_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS+
mean_acq_fix_count_high_iu_csm <-
mean(df$acq_csm_fix_count[df_long_acq_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS-
mean_acq_fix_count_low_iu_csm <-
mean(df$acq_csm_fix_count[df_long_acq_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS-
# combine into single variable called
all_mean_acq_fix_count <-
c(mean_acq_fix_count_high_iu_csp, mean_acq_fix_count_low_iu_csp,
mean_acq_fix_count_high_iu_csm, mean_acq_fix_count_low_iu_csm)
# obtain SD fix count for each group at each stimulus type and save as vector
sd_acq_fix_count_high_iu_csp <-
sd(df$acq_csp_fix_count[df_long_acq_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS+
sd_acq_fix_count_low_iu_csp <-
sd(df$acq_csp_fix_count[df_long_acq_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS+
sd_acq_fix_count_high_iu_csm <-
sd(df$acq_csm_fix_count[df_long_acq_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS-
sd_acq_fix_count_low_iu_csm <-
sd(df$acq_csm_fix_count[df_long_acq_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS-
# obtain SE:
se_acq_fix_count_high_iu_csp <- sd_acq_fix_count_high_iu_csp/sqrt(length(df$id))
se_acq_fix_count_low_iu_csp <- sd_acq_fix_count_low_iu_csp/sqrt(length(df$id))
se_acq_fix_count_high_iu_csm <- sd_acq_fix_count_high_iu_csm/sqrt(length(df$id))
se_acq_fix_count_low_iu_csm <- sd_acq_fix_count_low_iu_csm/sqrt(length(df$id))
# Combine all into single variable called all_se
all_se_acq_fix_count <- c(se_acq_fix_count_high_iu_csp, se_acq_fix_count_low_iu_csp,
se_acq_fix_count_high_iu_csm, se_acq_fix_count_low_iu_csm)
### Create new data frame for figures
# Which includes mean and SE for each condition
df_fig_acquisition_fix_count <- data.frame(all_mean_acq_fix_count, all_se_acq_fix_count)
### add labels
# add two more variables to indicate IU group and stimulus type.
# for IU group
df_fig_acquisition_fix_count$iu_group[1] <- "High IU"
df_fig_acquisition_fix_count$iu_group[2] <- "Low IU"
df_fig_acquisition_fix_count$iu_group[3] <- "High IU"
df_fig_acquisition_fix_count$iu_group[4] <- "Low IU"
# for stimulus
df_fig_acquisition_fix_count$stimulus[1] <- "CS+"
df_fig_acquisition_fix_count$stimulus[2] <- "CS+"
df_fig_acquisition_fix_count$stimulus[3] <- "CS-"
df_fig_acquisition_fix_count$stimulus[4] <- "CS-"
# and re-order levels of stimulus factor so that CS+ appears on left in the graph
df_fig_acquisition_fix_count$stimulus <-
factor(df_fig_acquisition_fix_count$stimulus,levels=c("CS+","CS-"))
### Create figure
fig_acquisition_fix_count <- ggplot(df_fig_acquisition_fix_count,
aes(x = iu_group, y = all_mean_acq_fix_count,
fill = stimulus)) +
geom_bar(stat = "identity", position = position_dodge(.6), width = .5, alpha = .85) +
scale_y_continuous(limits = c(0, 10), expand = c(0,0)) +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
theme(axis.text.y = element_text(size = 10),axis.ticks.y = element_line(size = 1),
axis.line.y = element_line(colour = "black")) +
theme(axis.text.x = element_text(colour = "black"), axis.ticks.x = element_blank(),
axis.line.x = element_line(colour = "black")) +
theme(legend.position = "bottom", legend.title = element_blank()) +
scale_fill_manual(values = c("#c45150", "#824372")) +
ggtitle("Acquisition") +
labs(y = "Mean Fixation Count", x = "Intolerance of Uncertainty") +
geom_errorbar(aes(ymin = all_mean_acq_fix_count - all_se_acq_fix_count,
ymax = all_mean_acq_fix_count + all_se_acq_fix_count),
width = .15, position = position_dodge(.6), colour = "#090707", size = .3)
print(fig_acquisition_fix_count) 
ggsave(filename = "graphs/bar_plots/acquisition_fix_count.png",
plot = fig_acquisition_fix_count,
width = 20,
height = 10,
dpi = 300,
units = "cm")Extinction
# obtain mean fix count for each group at each stimulus type and save as vector
mean_e_ext_fix_count_high_iu_csp <-
mean(df$e_ext_csp_fix_count[df_long_ext_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS+ early
mean_e_ext_fix_count_low_iu_csp <-
mean(df$e_ext_csp_fix_count[df_long_ext_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS+ early
mean_e_ext_fix_count_high_iu_csm <-
mean(df$e_ext_csm_fix_count[df_long_ext_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS- early
mean_e_ext_fix_count_low_iu_csm <-
mean(df$e_ext_csm_fix_count[df_long_ext_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS- early
mean_l_ext_fix_count_high_iu_csp <-
mean(df$l_ext_csp_fix_count[df_long_ext_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS+ late
mean_l_ext_fix_count_low_iu_csp <-
mean(df$l_ext_csp_fix_count[df_long_ext_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS+ late
mean_l_ext_fix_count_high_iu_csm <-
mean(df$l_ext_csm_fix_count[df_long_ext_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS- late
mean_l_ext_fix_count_low_iu_csm <-
mean(df$l_ext_csm_fix_count[df_long_ext_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS- late
# combine into single variable called
all_mean_ext_fix_count <-
c(mean_e_ext_fix_count_high_iu_csp, mean_e_ext_fix_count_low_iu_csp,
mean_e_ext_fix_count_high_iu_csm, mean_e_ext_fix_count_low_iu_csm,
mean_l_ext_fix_count_high_iu_csp, mean_l_ext_fix_count_low_iu_csp,
mean_l_ext_fix_count_high_iu_csm, mean_l_ext_fix_count_low_iu_csm)
# obtain SD fix count for each group at each stimulus type and save as vector
sd_e_ext_fix_count_high_iu_csp <-
sd(df$e_ext_csp_fix_count[df_long_ext_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS+ early
sd_e_ext_fix_count_low_iu_csp <-
sd(df$e_ext_csp_fix_count[df_long_ext_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS+ early
sd_e_ext_fix_count_high_iu_csm <-
sd(df$e_ext_csm_fix_count[df_long_ext_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS- early
sd_e_ext_fix_count_low_iu_csm <-
sd(df$e_ext_csm_fix_count[df_long_ext_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS- early
sd_l_ext_fix_count_high_iu_csp <-
sd(df$l_ext_csp_fix_count[df_long_ext_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS+ late
sd_l_ext_fix_count_low_iu_csp <-
sd(df$l_ext_csp_fix_count[df_long_ext_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS+ late
sd_l_ext_fix_count_high_iu_csm <-
sd(df$l_ext_csm_fix_count[df_long_ext_fix_count$iu_group == "1"], na.rm = TRUE) # high IU CS- late
sd_l_ext_fix_count_low_iu_csm <-
sd(df$l_ext_csm_fix_count[df_long_ext_fix_count$iu_group == "-1"], na.rm = TRUE) # low IU CS- late
# obtain SE:
se_e_ext_fix_count_high_iu_csp <- sd_e_ext_fix_count_high_iu_csp/sqrt(length(df$id))
se_e_ext_fix_count_low_iu_csp <- sd_e_ext_fix_count_low_iu_csp/sqrt(length(df$id))
se_e_ext_fix_count_high_iu_csm <- sd_e_ext_fix_count_high_iu_csm/sqrt(length(df$id))
se_e_ext_fix_count_low_iu_csm <- sd_e_ext_fix_count_low_iu_csm/sqrt(length(df$id))
se_l_ext_fix_count_high_iu_csp <- sd_l_ext_fix_count_high_iu_csp/sqrt(length(df$id))
se_l_ext_fix_count_low_iu_csp <- sd_l_ext_fix_count_low_iu_csp/sqrt(length(df$id))
se_l_ext_fix_count_high_iu_csm <- sd_l_ext_fix_count_high_iu_csm/sqrt(length(df$id))
se_l_ext_fix_count_low_iu_csm <- sd_l_ext_fix_count_low_iu_csm/sqrt(length(df$id))
# Combine all into single variable called all_se
all_se_ext_fix_count <- c(se_e_ext_fix_count_high_iu_csp, se_e_ext_fix_count_low_iu_csp,
se_e_ext_fix_count_high_iu_csm, se_e_ext_fix_count_low_iu_csm,
se_l_ext_fix_count_high_iu_csp, se_l_ext_fix_count_low_iu_csp,
se_l_ext_fix_count_high_iu_csm, se_l_ext_fix_count_low_iu_csm)
### Create new data frame for figures
# Which includes mean and SE for each condition
df_fig_extinction_fix_count <- data.frame(all_mean_ext_fix_count, all_se_ext_fix_count)
### add labels
# add two more variables to indicate IU group and stimulus type.
# for IU group
df_fig_extinction_fix_count$iu_group[1] <- "High IU"
df_fig_extinction_fix_count$iu_group[2] <- "Low IU"
df_fig_extinction_fix_count$iu_group[3] <- "High IU"
df_fig_extinction_fix_count$iu_group[4] <- "Low IU"
df_fig_extinction_fix_count$iu_group[5] <- "High IU"
df_fig_extinction_fix_count$iu_group[6] <- "Low IU"
df_fig_extinction_fix_count$iu_group[7] <- "High IU"
df_fig_extinction_fix_count$iu_group[8] <- "Low IU"
# for stimulus
df_fig_extinction_fix_count$stimulus[1] <- "CS+"
df_fig_extinction_fix_count$stimulus[2] <- "CS+"
df_fig_extinction_fix_count$stimulus[3] <- "CS-"
df_fig_extinction_fix_count$stimulus[4] <- "CS-"
df_fig_extinction_fix_count$stimulus[5] <- "CS+"
df_fig_extinction_fix_count$stimulus[6] <- "CS+"
df_fig_extinction_fix_count$stimulus[7] <- "CS-"
df_fig_extinction_fix_count$stimulus[8] <- "CS-"
# and re-order levels of stimulus factor so that CS+ appears on left in the graph
df_fig_extinction_fix_count$stimulus <-
factor(df_fig_extinction_fix_count$stimulus,levels=c("CS+","CS-"))
# for early / late extinction
df_fig_extinction_fix_count$time[1] <- "Early"
df_fig_extinction_fix_count$time[2] <- "Early"
df_fig_extinction_fix_count$time[3] <- "Early"
df_fig_extinction_fix_count$time[4] <- "Early"
df_fig_extinction_fix_count$time[5] <- "Late"
df_fig_extinction_fix_count$time[6] <- "Late"
df_fig_extinction_fix_count$time[7] <- "Late"
df_fig_extinction_fix_count$time[8] <- "Late"
### create figure
fig_extinction_fix_count <- ggplot(df_fig_extinction_fix_count,
aes(x = iu_group, y = all_mean_ext_fix_count,
fill = stimulus)) +
geom_bar(stat = "identity", position = position_dodge(.6), width = .5, alpha = .85) +
scale_y_continuous(limits = c(0, 10), expand = c(0,0)) +
facet_wrap(~ time) +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
theme(axis.text.y = element_text(size = 10),axis.ticks.y = element_line(size = 1),
axis.line.y = element_line(colour = "black")) +
theme(axis.text.x = element_text(colour = "black"), axis.ticks.x = element_blank(),
axis.line.x = element_line(colour = "black")) +
theme(legend.position = "bottom", legend.title = element_blank()) +
scale_fill_manual(values = c("#c45150", "#824372")) +
ggtitle("Extinction") +
labs(y = "Mean Fixation Count", x = "Intolerance of Uncertainty") +
geom_errorbar(aes(ymin = all_mean_ext_fix_count - all_se_ext_fix_count,
ymax = all_mean_ext_fix_count + all_se_ext_fix_count),
width = .15, position = position_dodge(.6), colour = "#090707", size = .3) +
theme(strip.background = element_blank()) +
theme(strip.text = element_text(size = 12))
# obtain and check figure
print(fig_extinction_fix_count) 
# save figure to files
ggsave(filename = "graphs/bar_plots/extinction_fix_count.png",
plot = fig_extinction_fix_count,
width = 20,
height = 10,
dpi = 300,
units = "cm")Fixation Duration (Log Transformed)
Acquisition
# obtain mean fix duration for each group at each stimulus type and save as vector
mean_acq_fix_duration_high_iu_csp_log <-
mean(df$acq_csp_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "1"], na.rm = TRUE) # high IU CS+
mean_acq_fix_duration_low_iu_csp_log <-
mean(df$acq_csp_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "-1"], na.rm = TRUE) # low IU CS+
mean_acq_fix_duration_high_iu_csm_log <-
mean(df$acq_csm_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "1"], na.rm = TRUE) # high IU CS-
mean_acq_fix_duration_low_iu_csm_log <-
mean(df$acq_csm_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "-1"], na.rm = TRUE) # low IU CS-
# combine into single variable called
all_mean_acq_fix_duration_log <-
c(mean_acq_fix_duration_high_iu_csp_log, mean_acq_fix_duration_low_iu_csp_log,
mean_acq_fix_duration_high_iu_csm_log, mean_acq_fix_duration_low_iu_csm_log)
# obtain SD fix duration for each group at each stimulus type and save as vector
sd_acq_fix_duration_high_iu_csp_log <-
sd(df$acq_csp_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "1"], na.rm = TRUE) # high IU CS+
sd_acq_fix_duration_low_iu_csp_log <-
sd(df$acq_csp_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "-1"], na.rm = TRUE) # low IU CS+
sd_acq_fix_duration_high_iu_csm_log <-
sd(df$acq_csm_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "1"], na.rm = TRUE) # high IU CS-
sd_acq_fix_duration_low_iu_csm_log <-
sd(df$acq_csm_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "-1"], na.rm = TRUE) # low IU CS-
# obtain SE:
se_acq_fix_duration_high_iu_csp_log <- sd_acq_fix_duration_high_iu_csp_log/sqrt(length(df$id))
se_acq_fix_duration_low_iu_csp_log <- sd_acq_fix_duration_low_iu_csp_log/sqrt(length(df$id))
se_acq_fix_duration_high_iu_csm_log <- sd_acq_fix_duration_high_iu_csm_log/sqrt(length(df$id))
se_acq_fix_duration_low_iu_csm_log <- sd_acq_fix_duration_low_iu_csm_log/sqrt(length(df$id))
# combine all into single variable called all_se
all_se_acq_fix_duration_log <- c(se_acq_fix_duration_high_iu_csp_log, se_acq_fix_duration_low_iu_csp_log,
se_acq_fix_duration_high_iu_csm_log, se_acq_fix_duration_low_iu_csm_log)
# create new data frame for figures, which includes mean and se
# for each condition
df_fig_acquisition_fix_duration_log <- data.frame(all_mean_acq_fix_duration_log, all_se_acq_fix_duration_log)
# add labels - add two more variables to indicate IU group and stimulus type.
# for IU group
df_fig_acquisition_fix_duration_log$iu_group[1] <- "High IU"
df_fig_acquisition_fix_duration_log$iu_group[2] <- "Low IU"
df_fig_acquisition_fix_duration_log$iu_group[3] <- "High IU"
df_fig_acquisition_fix_duration_log$iu_group[4] <- "Low IU"
# for stimulus
df_fig_acquisition_fix_duration_log$stimulus[1] <- "CS+"
df_fig_acquisition_fix_duration_log$stimulus[2] <- "CS+"
df_fig_acquisition_fix_duration_log$stimulus[3] <- "CS-"
df_fig_acquisition_fix_duration_log$stimulus[4] <- "CS-"
# and re-order levels of stimulus factor so that CS+ appears on left in the graph
df_fig_acquisition_fix_duration_log$stimulus <-
factor(df_fig_acquisition_fix_duration_log$stimulus,levels = c("CS+","CS-"))
# create figure
fig_acquisition_fix_duration_log <- ggplot(df_fig_acquisition_fix_duration_log,
aes(x = iu_group, y = all_mean_acq_fix_duration_log,
fill = stimulus)) +
geom_bar(stat = "identity", position = position_dodge(.6), width = .5, alpha = .85) +
scale_y_continuous(limits = c(0, 7.2), expand = c(0,0)) +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
theme(axis.text.y = element_text(size = 10),axis.ticks.y = element_line(size = 1),
axis.line.y = element_line(colour = "black")) +
theme(axis.text.x = element_text(colour = "black"), axis.ticks.x = element_blank(),
axis.line.x = element_line(colour = "black")) +
theme(legend.position = "bottom", legend.title = element_blank()) +
scale_fill_manual(values = c("#c45150", "#824372")) +
ggtitle("Acquisition") +
labs(y = "Mean Fixation Duration (ms) (Log Transformed)", x = "Intolerance of Uncertainty") +
geom_errorbar(aes(ymin = all_mean_acq_fix_duration_log - all_se_acq_fix_duration_log,
ymax = all_mean_acq_fix_duration_log + all_se_acq_fix_duration_log),
width = .15, position = position_dodge(.6), colour = "#090707", size = .3)
# obtain and check figure
print(fig_acquisition_fix_duration_log) 
# save figure to file
ggsave(filename = "graphs/bar_plots/acquisition_fix_duration_log.png",
plot = fig_acquisition_fix_duration_log,
width = 20,
height = 10,
dpi = 300,
units = "cm")Extinction
# obtain mean fix duration for each group at each stimulus type and save as vector
# high IU CS+ early
mean_e_ext_fix_duration_high_iu_csp_log <-
mean(df$e_ext_csp_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "1"], na.rm = TRUE)
# low IU CS+ early
mean_e_ext_fix_duration_low_iu_csp_log <-
mean(df$e_ext_csp_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "-1"], na.rm = TRUE)
# high IU CS- early
mean_e_ext_fix_duration_high_iu_csm_log <-
mean(df$e_ext_csm_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "1"], na.rm = TRUE)
# low IU CS- early
mean_e_ext_fix_duration_low_iu_csm_log <-
mean(df$e_ext_csm_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "-1"], na.rm = TRUE)
# high IU CS+ late
mean_l_ext_fix_duration_high_iu_csp_log <-
mean(df$l_ext_csp_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "1"], na.rm = TRUE)
# low IU CS+ late
mean_l_ext_fix_duration_low_iu_csp_log <-
mean(df$l_ext_csp_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "-1"], na.rm = TRUE)
# high IU CS- late
mean_l_ext_fix_duration_high_iu_csm_log <-
mean(df$l_ext_csm_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "1"], na.rm = TRUE)
# low IU CS- late
mean_l_ext_fix_duration_low_iu_csm_log <-
mean(df$l_ext_csm_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "-1"], na.rm = TRUE)
# combine into single variable called
all_mean_ext_fix_duration_log <-
c(mean_e_ext_fix_duration_high_iu_csp_log, mean_e_ext_fix_duration_low_iu_csp_log,
mean_e_ext_fix_duration_high_iu_csm_log, mean_e_ext_fix_duration_low_iu_csm_log,
mean_l_ext_fix_duration_high_iu_csp_log, mean_l_ext_fix_duration_low_iu_csp_log,
mean_l_ext_fix_duration_high_iu_csm_log, mean_l_ext_fix_duration_low_iu_csm_log)
# obtain SD fix duration for each group at each stimulus type and save as vector
# high IU CS+ early
sd_e_ext_fix_duration_high_iu_csp_log <-
sd(df$e_ext_csp_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "1"], na.rm = TRUE)
# low IU CS+ early
sd_e_ext_fix_duration_low_iu_csp_log <-
sd(df$e_ext_csp_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "-1"], na.rm = TRUE)
# high IU CS- early
sd_e_ext_fix_duration_high_iu_csm_log <-
sd(df$e_ext_csm_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "1"], na.rm = TRUE)
# low IU CS- early
sd_e_ext_fix_duration_low_iu_csm_log <-
sd(df$e_ext_csm_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "-1"], na.rm = TRUE)
# high IU CS+ late
sd_l_ext_fix_duration_high_iu_csp_log <-
sd(df$l_ext_csp_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "1"], na.rm = TRUE)
# low IU CS+ late
sd_l_ext_fix_duration_low_iu_csp_log <-
sd(df$l_ext_csp_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "-1"], na.rm = TRUE)
# high IU CS- late
sd_l_ext_fix_duration_high_iu_csm_log <-
sd(df$l_ext_csm_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "1"], na.rm = TRUE)
# low IU CS- late
sd_l_ext_fix_duration_low_iu_csm_log <-
sd(df$l_ext_csm_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "-1"], na.rm = TRUE)
# obtain SE:
se_e_ext_fix_duration_high_iu_csp_log <- sd_e_ext_fix_duration_high_iu_csp_log/sqrt(length(df$id))
se_e_ext_fix_duration_low_iu_csp_log <- sd_e_ext_fix_duration_low_iu_csp_log/sqrt(length(df$id))
se_e_ext_fix_duration_high_iu_csm_log <- sd_e_ext_fix_duration_high_iu_csm_log/sqrt(length(df$id))
se_e_ext_fix_duration_low_iu_csm_log <- sd_e_ext_fix_duration_low_iu_csm_log/sqrt(length(df$id))
se_l_ext_fix_duration_high_iu_csp_log <- sd_l_ext_fix_duration_high_iu_csp_log/sqrt(length(df$id))
se_l_ext_fix_duration_low_iu_csp_log <- sd_l_ext_fix_duration_low_iu_csp_log/sqrt(length(df$id))
se_l_ext_fix_duration_high_iu_csm_log <- sd_l_ext_fix_duration_high_iu_csm_log/sqrt(length(df$id))
se_l_ext_fix_duration_low_iu_csm_log <- sd_l_ext_fix_duration_low_iu_csm_log/sqrt(length(df$id))
# combine all into single variable
all_se_ext_fix_duration_log <- c(se_e_ext_fix_duration_high_iu_csp_log, se_e_ext_fix_duration_low_iu_csp_log,
se_e_ext_fix_duration_high_iu_csm_log, se_e_ext_fix_duration_low_iu_csm_log,
se_l_ext_fix_duration_high_iu_csp_log, se_l_ext_fix_duration_low_iu_csp_log,
se_l_ext_fix_duration_high_iu_csm_log, se_l_ext_fix_duration_low_iu_csm_log)
# create new data frame for figures which includes mean and SE for each condition
df_fig_extinction_fix_duration_log <- data.frame(all_mean_ext_fix_duration_log, all_se_ext_fix_duration_log)
# add labels - add two more variables to indicate IU group, stimulus type and extinction time
# for IU group
df_fig_extinction_fix_duration_log$iu_group[1] <- "High IU"
df_fig_extinction_fix_duration_log$iu_group[2] <- "Low IU"
df_fig_extinction_fix_duration_log$iu_group[3] <- "High IU"
df_fig_extinction_fix_duration_log$iu_group[4] <- "Low IU"
df_fig_extinction_fix_duration_log$iu_group[5] <- "High IU"
df_fig_extinction_fix_duration_log$iu_group[6] <- "Low IU"
df_fig_extinction_fix_duration_log$iu_group[7] <- "High IU"
df_fig_extinction_fix_duration_log$iu_group[8] <- "Low IU"
# for stimulus
df_fig_extinction_fix_duration_log$stimulus[1] <- "CS+"
df_fig_extinction_fix_duration_log$stimulus[2] <- "CS+"
df_fig_extinction_fix_duration_log$stimulus[3] <- "CS-"
df_fig_extinction_fix_duration_log$stimulus[4] <- "CS-"
df_fig_extinction_fix_duration_log$stimulus[5] <- "CS+"
df_fig_extinction_fix_duration_log$stimulus[6] <- "CS+"
df_fig_extinction_fix_duration_log$stimulus[7] <- "CS-"
df_fig_extinction_fix_duration_log$stimulus[8] <- "CS-"
# and re-order levels of stimulus factor so that CS+ appears on left in the graph
df_fig_extinction_fix_duration_log$stimulus <-
factor(df_fig_extinction_fix_duration_log$stimulus,levels=c("CS+","CS-"))
# for early / late extinction
df_fig_extinction_fix_duration_log$time[1] <- "Early"
df_fig_extinction_fix_duration_log$time[2] <- "Early"
df_fig_extinction_fix_duration_log$time[3] <- "Early"
df_fig_extinction_fix_duration_log$time[4] <- "Early"
df_fig_extinction_fix_duration_log$time[5] <- "Late"
df_fig_extinction_fix_duration_log$time[6] <- "Late"
df_fig_extinction_fix_duration_log$time[7] <- "Late"
df_fig_extinction_fix_duration_log$time[8] <- "Late"
# create figure
fig_extinction_fix_duration_log <- ggplot(df_fig_extinction_fix_duration_log,
aes(x = iu_group, y = all_mean_ext_fix_duration_log,
fill = stimulus)) +
geom_bar(stat = "identity", position = position_dodge(.6), width = .5, alpha = .85) +
scale_y_continuous(limits = c(0, 7.2), expand = c(0,0)) +
facet_wrap(~ time) +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
theme(axis.text.y = element_text(size = 10),axis.ticks.y = element_line(size = 1),
axis.line.y = element_line(colour = "black")) +
theme(axis.text.x = element_text(colour = "black"), axis.ticks.x = element_blank(),
axis.line.x = element_line(colour = "black")) +
theme(legend.position = "bottom", legend.title = element_blank()) +
scale_fill_manual(values = c("#c45150", "#824372")) +
ggtitle("Extinction") +
labs(y = "Mean Fixation Duration (ms) (Log Transformed)", x = "Intolerance of Uncertainty") +
geom_errorbar(aes(ymin = all_mean_ext_fix_duration_log - all_se_ext_fix_duration_log,
ymax = all_mean_ext_fix_duration_log + all_se_ext_fix_duration_log),
width = .15, position = position_dodge(.6), colour = "#090707", size = .3) +
theme(strip.background = element_blank()) +
theme(strip.text = element_text(size = 12))
# obtain and check figure
print(fig_extinction_fix_duration_log) 
# save figure to files
ggsave(filename = "graphs/bar_plots/extinction_fix_duration_log.png",
plot = fig_extinction_fix_duration_log,
width = 20,
height = 10,
dpi = 300,
units = "cm")Saccade Amplitude
Acquisition
# obtain mean sacc amplitude for each group at each stimulus type and save as vector
mean_acq_sacc_amplitude_high_iu_csp <-
mean(df$acq_csp_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "1"], na.rm = TRUE) # high IU CS+
mean_acq_sacc_amplitude_low_iu_csp <-
mean(df$acq_csp_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "-1"], na.rm = TRUE) # low IU CS+
mean_acq_sacc_amplitude_high_iu_csm <-
mean(df$acq_csm_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "1"], na.rm = TRUE) # high IU CS-
mean_acq_sacc_amplitude_low_iu_csm <-
mean(df$acq_csm_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "-1"], na.rm = TRUE) # low IU CS-
# combine into single variable called
all_mean_acq_sacc_amplitude <-
c(mean_acq_sacc_amplitude_high_iu_csp, mean_acq_sacc_amplitude_low_iu_csp,
mean_acq_sacc_amplitude_high_iu_csm, mean_acq_sacc_amplitude_low_iu_csm)
# obtain SD sacc amplitude for each group at each stimulus type and save as vector
sd_acq_sacc_amplitude_high_iu_csp <-
sd(df$acq_csp_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "1"], na.rm = TRUE) # high IU CS+
sd_acq_sacc_amplitude_low_iu_csp <-
sd(df$acq_csp_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "-1"], na.rm = TRUE) # low IU CS+
sd_acq_sacc_amplitude_high_iu_csm <-
sd(df$acq_csm_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "1"], na.rm = TRUE) # high IU CS-
sd_acq_sacc_amplitude_low_iu_csm <-
sd(df$acq_csm_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "-1"], na.rm = TRUE) # low IU CS-
# obtain SE:
se_acq_sacc_amplitude_high_iu_csp <- sd_acq_sacc_amplitude_high_iu_csp/sqrt(length(df$id))
se_acq_sacc_amplitude_low_iu_csp <- sd_acq_sacc_amplitude_low_iu_csp/sqrt(length(df$id))
se_acq_sacc_amplitude_high_iu_csm <- sd_acq_sacc_amplitude_high_iu_csm/sqrt(length(df$id))
se_acq_sacc_amplitude_low_iu_csm <- sd_acq_sacc_amplitude_low_iu_csm/sqrt(length(df$id))
# combine all into single variable called all_se
all_se_acq_sacc_amplitude <- c(se_acq_sacc_amplitude_high_iu_csp, se_acq_sacc_amplitude_low_iu_csp,
se_acq_sacc_amplitude_high_iu_csm, se_acq_sacc_amplitude_low_iu_csm)
# create new data frame for figures, which includes mean and se
# for each condition
df_fig_acquisition_sacc_amplitude <- data.frame(all_mean_acq_sacc_amplitude, all_se_acq_sacc_amplitude)
# add labels - add two more variables to indicate IU group and stimulus type.
# for IU group
df_fig_acquisition_sacc_amplitude$iu_group[1] <- "High IU"
df_fig_acquisition_sacc_amplitude$iu_group[2] <- "Low IU"
df_fig_acquisition_sacc_amplitude$iu_group[3] <- "High IU"
df_fig_acquisition_sacc_amplitude$iu_group[4] <- "Low IU"
# for stimulus
df_fig_acquisition_sacc_amplitude$stimulus[1] <- "CS+"
df_fig_acquisition_sacc_amplitude$stimulus[2] <- "CS+"
df_fig_acquisition_sacc_amplitude$stimulus[3] <- "CS-"
df_fig_acquisition_sacc_amplitude$stimulus[4] <- "CS-"
# and re-order levels of stimulus factor so that CS+ appears on left in the graph
df_fig_acquisition_sacc_amplitude$stimulus <-
factor(df_fig_acquisition_sacc_amplitude$stimulus,levels=c("CS+","CS-"))
# create figure
fig_acquisition_sacc_amplitude <- ggplot(df_fig_acquisition_sacc_amplitude,
aes(x = iu_group, y = all_mean_acq_sacc_amplitude,
fill = stimulus)) +
geom_bar(stat = "identity", position = position_dodge(.6), width = .5, alpha = .85) +
scale_y_continuous(limits = c(0, 4), expand = c(0,0)) +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
theme(axis.text.y = element_text(size = 10),axis.ticks.y = element_line(size = 1),
axis.line.y = element_line(colour = "black")) +
theme(axis.text.x = element_text(colour = "black"), axis.ticks.x = element_blank(),
axis.line.x = element_line(colour = "black")) +
theme(legend.position = "bottom", legend.title = element_blank()) +
scale_fill_manual(values = c("#c45150", "#824372")) +
ggtitle("Acquisition") +
labs(y = "Mean Saccade Amplitude (degrees/ms)", x = "Intolerance of Uncertainty") +
geom_errorbar(aes(ymin = all_mean_acq_sacc_amplitude - all_se_acq_sacc_amplitude,
ymax = all_mean_acq_sacc_amplitude + all_se_acq_sacc_amplitude),
width = .15, position = position_dodge(.6), colour = "#090707", size = .3)
# obtain and check figure
print(fig_acquisition_sacc_amplitude) 
# save figure to file
ggsave(filename = "graphs/bar_plots/acquisition_sacc_amplitude.png",
plot = fig_acquisition_sacc_amplitude,
width = 20,
height = 10,
dpi = 300,
units = "cm")Extinction
# obtain mean sacc amplitude for each group at each stimulus type and save as vector
# high IU CS+ early
mean_e_ext_sacc_amplitude_high_iu_csp <-
mean(df$e_ext_csp_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "1"], na.rm = TRUE)
# low IU CS+ early
mean_e_ext_sacc_amplitude_low_iu_csp <-
mean(df$e_ext_csp_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "-1"], na.rm = TRUE)
# high IU CS- early
mean_e_ext_sacc_amplitude_high_iu_csm <-
mean(df$e_ext_csm_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "1"], na.rm = TRUE)
# low IU CS- early
mean_e_ext_sacc_amplitude_low_iu_csm <-
mean(df$e_ext_csm_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "-1"], na.rm = TRUE)
# high IU CS+ late
mean_l_ext_sacc_amplitude_high_iu_csp <-
mean(df$l_ext_csp_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "1"], na.rm = TRUE)
# low IU CS+ late
mean_l_ext_sacc_amplitude_low_iu_csp <-
mean(df$l_ext_csp_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "-1"], na.rm = TRUE)
# high IU CS- late
mean_l_ext_sacc_amplitude_high_iu_csm <-
mean(df$l_ext_csm_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "1"], na.rm = TRUE)
# low IU CS- late
mean_l_ext_sacc_amplitude_low_iu_csm <-
mean(df$l_ext_csm_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "-1"], na.rm = TRUE)
# combine into single variable called
all_mean_ext_sacc_amplitude <-
c(mean_e_ext_sacc_amplitude_high_iu_csp, mean_e_ext_sacc_amplitude_low_iu_csp,
mean_e_ext_sacc_amplitude_high_iu_csm, mean_e_ext_sacc_amplitude_low_iu_csm,
mean_l_ext_sacc_amplitude_high_iu_csp, mean_l_ext_sacc_amplitude_low_iu_csp,
mean_l_ext_sacc_amplitude_high_iu_csm, mean_l_ext_sacc_amplitude_low_iu_csm)
# obtain SD sacc amplitude for each group at each stimulus type and save as vector
# high IU CS+ early
sd_e_ext_sacc_amplitude_high_iu_csp <-
sd(df$e_ext_csp_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "1"], na.rm = TRUE)
# low IU CS+ early
sd_e_ext_sacc_amplitude_low_iu_csp <-
sd(df$e_ext_csp_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "-1"], na.rm = TRUE)
# high IU CS- early
sd_e_ext_sacc_amplitude_high_iu_csm <-
sd(df$e_ext_csm_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "1"], na.rm = TRUE)
# low IU CS- early
sd_e_ext_sacc_amplitude_low_iu_csm <-
sd(df$e_ext_csm_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "-1"], na.rm = TRUE)
# high IU CS+ late
sd_l_ext_sacc_amplitude_high_iu_csp <-
sd(df$l_ext_csp_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "1"], na.rm = TRUE)
# low IU CS+ late
sd_l_ext_sacc_amplitude_low_iu_csp <-
sd(df$l_ext_csp_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "-1"], na.rm = TRUE)
# high IU CS- late
sd_l_ext_sacc_amplitude_high_iu_csm <-
sd(df$l_ext_csm_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "1"], na.rm = TRUE)
# low IU CS- late
sd_l_ext_sacc_amplitude_low_iu_csm <-
sd(df$l_ext_csm_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "-1"], na.rm = TRUE)
# obtain SE:
se_e_ext_sacc_amplitude_high_iu_csp <- sd_e_ext_sacc_amplitude_high_iu_csp/sqrt(length(df$id))
se_e_ext_sacc_amplitude_low_iu_csp <- sd_e_ext_sacc_amplitude_low_iu_csp/sqrt(length(df$id))
se_e_ext_sacc_amplitude_high_iu_csm <- sd_e_ext_sacc_amplitude_high_iu_csm/sqrt(length(df$id))
se_e_ext_sacc_amplitude_low_iu_csm <- sd_e_ext_sacc_amplitude_low_iu_csm/sqrt(length(df$id))
se_l_ext_sacc_amplitude_high_iu_csp <- sd_l_ext_sacc_amplitude_high_iu_csp/sqrt(length(df$id))
se_l_ext_sacc_amplitude_low_iu_csp <- sd_l_ext_sacc_amplitude_low_iu_csp/sqrt(length(df$id))
se_l_ext_sacc_amplitude_high_iu_csm <- sd_l_ext_sacc_amplitude_high_iu_csm/sqrt(length(df$id))
se_l_ext_sacc_amplitude_low_iu_csm <- sd_l_ext_sacc_amplitude_low_iu_csm/sqrt(length(df$id))
# combine all into single variable
all_se_ext_sacc_amplitude <- c(se_e_ext_sacc_amplitude_high_iu_csp, se_e_ext_sacc_amplitude_low_iu_csp,
se_e_ext_sacc_amplitude_high_iu_csm, se_e_ext_sacc_amplitude_low_iu_csm,
se_l_ext_sacc_amplitude_high_iu_csp, se_l_ext_sacc_amplitude_low_iu_csp,
se_l_ext_sacc_amplitude_high_iu_csm, se_l_ext_sacc_amplitude_low_iu_csm)
# create new data frame for figures which includes mean and SE for each condition
df_fig_extinction_sacc_amplitude <- data.frame(all_mean_ext_sacc_amplitude, all_se_ext_sacc_amplitude)
# add labels - add two more variables to indicate IU group, stimulus type and extinction time
# for IU group
df_fig_extinction_sacc_amplitude$iu_group[1] <- "High IU"
df_fig_extinction_sacc_amplitude$iu_group[2] <- "Low IU"
df_fig_extinction_sacc_amplitude$iu_group[3] <- "High IU"
df_fig_extinction_sacc_amplitude$iu_group[4] <- "Low IU"
df_fig_extinction_sacc_amplitude$iu_group[5] <- "High IU"
df_fig_extinction_sacc_amplitude$iu_group[6] <- "Low IU"
df_fig_extinction_sacc_amplitude$iu_group[7] <- "High IU"
df_fig_extinction_sacc_amplitude$iu_group[8] <- "Low IU"
# for stimulus
df_fig_extinction_sacc_amplitude$stimulus[1] <- "CS+"
df_fig_extinction_sacc_amplitude$stimulus[2] <- "CS+"
df_fig_extinction_sacc_amplitude$stimulus[3] <- "CS-"
df_fig_extinction_sacc_amplitude$stimulus[4] <- "CS-"
df_fig_extinction_sacc_amplitude$stimulus[5] <- "CS+"
df_fig_extinction_sacc_amplitude$stimulus[6] <- "CS+"
df_fig_extinction_sacc_amplitude$stimulus[7] <- "CS-"
df_fig_extinction_sacc_amplitude$stimulus[8] <- "CS-"
# and re-order levels of stimulus factor so that CS+ appears on left in the graph
df_fig_extinction_sacc_amplitude$stimulus <-
factor(df_fig_extinction_sacc_amplitude$stimulus,levels=c("CS+","CS-"))
# for early / late extinction
df_fig_extinction_sacc_amplitude$time[1] <- "Early"
df_fig_extinction_sacc_amplitude$time[2] <- "Early"
df_fig_extinction_sacc_amplitude$time[3] <- "Early"
df_fig_extinction_sacc_amplitude$time[4] <- "Early"
df_fig_extinction_sacc_amplitude$time[5] <- "Late"
df_fig_extinction_sacc_amplitude$time[6] <- "Late"
df_fig_extinction_sacc_amplitude$time[7] <- "Late"
df_fig_extinction_sacc_amplitude$time[8] <- "Late"
# create figure
fig_extinction_sacc_amplitude <- ggplot(df_fig_extinction_sacc_amplitude,
aes(x = iu_group, y = all_mean_ext_sacc_amplitude,
fill = stimulus)) +
geom_bar(stat = "identity", position = position_dodge(.6), width = .5, alpha = .85) +
scale_y_continuous(limits = c(0, 4), expand = c(0,0)) +
facet_wrap(~ time) +
theme_classic() +
theme(text = element_text(family = "serif"),
plot.title = element_text(face = "bold", hjust = 0.5, size = 15)) +
theme(axis.text.y = element_text(size = 10),axis.ticks.y = element_line(size = 1),
axis.line.y = element_line(colour = "black")) +
theme(axis.text.x = element_text(colour = "black"), axis.ticks.x = element_blank(),
axis.line.x = element_line(colour = "black")) +
theme(legend.position = "bottom", legend.title = element_blank()) +
scale_fill_manual(values = c("#c45150", "#824372")) +
ggtitle("Extinction") +
labs(y = "Mean Saccade Amplitude (degrees/ms)", x = "Intolerance of Uncertainty") +
geom_errorbar(aes(ymin = all_mean_ext_sacc_amplitude - all_se_ext_sacc_amplitude,
ymax = all_mean_ext_sacc_amplitude + all_se_ext_sacc_amplitude),
width = .15, position = position_dodge(.6), colour = "#090707", size = .3) +
theme(strip.background = element_blank()) +
theme(strip.text = element_text(size = 12))
# obtain and check figure
print(fig_extinction_sacc_amplitude) 
# save figure to files
ggsave(filename = "graphs/bar_plots/extinction_sacc_amplitude.png",
plot = fig_extinction_sacc_amplitude,
width = 20,
height = 10,
dpi = 300,
units = "cm")Combine Bar Plots
all_bar_plots <- grid.arrange(fig_acquisition_fix_count, fig_extinction_fix_count,
fig_acquisition_fix_duration_log, fig_extinction_fix_duration_log,
fig_acquisition_sacc_amplitude, fig_extinction_sacc_amplitude,
ncol = 2)
# save figure to files
ggsave(filename = "graphs/bar_plots/all_bar_plots.png",
plot = all_bar_plots,
width = 40,
height = 30,
dpi = 300,
units = "cm")ANCOVAs to test Specificity of IU over Trait Anxiety
ANCOVA Acquisition Fixation Count
# transform wide format data into long format for mixed ANCOVA
df_long_acq_fix_count <- melt(df, id = c("id", "iu_group", "sticsa_total"),
measure.vars = c("acq_csp_fix_count",
"acq_csm_fix_count"))
# rename columns for easier interpretation
colnames(df_long_acq_fix_count) = c("id", "iu_group", "sticsa_total", "condition", "fix_count")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_acq_fix_count$stimulus <-
factor(ifelse(df_long_acq_fix_count$condition == "acq_csp_fix_count", 1, -1))
# mean centre continuous covariate (STICSA)
# to apply mean centring, first obtain average sticsa scores for all participants,
# and save as a variable
df_long_acq_fix_count$sticsa_total_avg <- mean(df_long_acq_fix_count$sticsa_total)
# next, subtract this average from all participants' sticsa scores,
# and save as a variable
df_long_acq_fix_count$sticsa_total_centred <-
df_long_acq_fix_count$sticsa_total - df_long_acq_fix_count$sticsa_total_avg
# from this we have mean sticsa scores after centring
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) mixed ANCOVA,
# with mean-cenred STICSA as covariate
# and obtain effect size (partial eta squared)
acq_fix_count_ancova <-
anova_test(df_long_acq_fix_count, fix_count ~ iu_group * stimulus + Error(id/stimulus),
covariate = sticsa_total_centred, effect.size = "pes")
# obtain the mixed ANCOVA results
get_anova_table(acq_fix_count_ancova)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 sticsa_total_centred 1 136 0.059 0.808000 0.000434
## 2 iu_group 1 136 3.191 0.076000 0.023000
## 3 stimulus 1 136 11.622 0.000858 * 0.079000
## 4 sticsa_total_centred:stimulus 1 136 1.845 0.177000 0.013000
## 5 iu_group:stimulus 1 136 1.230 0.269000 0.009000# results:
# STICSA (centred): F(1,136) = 0.06, p = .808, eta2(partial) = < .001
# IU: F(1,136) = 3.19, p = .076, eta2(partial) = .023
# Stimulus: F(1,136) = 11.62, p < .001***, eta2(partial) = .079
# STICSA * Stimulus: F(1,136) = 1.85, p = .177, eta2(partial) = .013
# IU * Stimulus: F(1, 136) = 1.23, p = .269, eta2(partial) = .009
# therefore, after accounting for trait anxiety, IU no longer has a significant
# effect on fixation count in acquisition, but stimulus continues to have
# significant effect. IU*Stimulus interaction also remains non-significant,
# even after controlling for trait anxiety.
# write to csv
write.csv((get_anova_table(acq_fix_count_ancova)),
file = "tables/ancovas/acq_fix_count_ancova.csv")
# as there was a significant main effect of stimulus, obtain estimated
# marginal means to be reported (temporarily from SPSS):
emmeans_acq_fix_count_ancova_csp <- 6.92
emmeans_acq_fix_count_ancova_csm <- 7.32ANCOVA Acquisition Fixation Duration (Log Transformed)
# transform wide format data into long format for mixed ANCOVA
df_long_acq_fix_duration_log <- melt(df, id = c("id", "iu_group", "sticsa_total"),
measure.vars = c("acq_csp_fix_duration_log",
"acq_csm_fix_duration_log"))
# rename columns for easier interpretation
colnames(df_long_acq_fix_duration_log) = c("id", "iu_group", "sticsa_total", "condition", "fix_duration_log")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_acq_fix_duration_log$stimulus <-
factor(ifelse(df_long_acq_fix_duration_log$condition == "acq_csp_fix_duration_log", 1, -1))
# mean centre continuous covariate (STICSA)
# to apply mean centring, first obtain average sticsa scores for all participants,
# and save as a variable
df_long_acq_fix_duration_log$sticsa_total_avg <- mean(df_long_acq_fix_duration_log$sticsa_total)
# next, subtract this average from all participants' sticsa scores,
# and save as a variable
df_long_acq_fix_duration_log$sticsa_total_centred <-
df_long_acq_fix_duration_log$sticsa_total - df_long_acq_fix_duration_log$sticsa_total_avg
# from this we have mean sticsa scores after centring
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) mixed ANCOVA,
# with mean-cenred STICSA as covariate
# and obtain effect size (partial eta squared)
acq_fix_duration_ancova_log <-
anova_test(df_long_acq_fix_duration_log, fix_duration_log ~ iu_group * stimulus + Error(id/stimulus),
covariate = sticsa_total_centred, effect.size = "pes")
# obtain the mixed ANCOVA results
get_anova_table(acq_fix_duration_ancova_log)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 sticsa_total_centred 1 136 0.268 0.606 0.002
## 2 iu_group 1 136 3.890 0.051 0.028
## 3 stimulus 1 136 2.935 0.089 0.021
## 4 sticsa_total_centred:stimulus 1 136 0.409 0.524 0.003
## 5 iu_group:stimulus 1 136 1.674 0.198 0.012# results:
# STICSA (centred): F(1,136) = 0.27, p = .606, eta2(partial) = .002
# IU: F(1,136) = 3.89, p = .051, eta2(partial) = .028
# Stimulus: F(1,136) = 2.94, p = .089, eta2(partial) = .021
# STICSA * Stimulus: F(1,136) = 0.41, p = .524, eta2(partial) = .003
# IU * Stimulus: F(1, 136) = 1.67, p = .198, eta2(partial) = .012
# there are no significant effects or interactions on fixation duration in acquisition.
# write to csv
write.csv((get_anova_table(acq_fix_duration_ancova_log)),
file = "tables/ancovas/acq_fix_duration_ancova_log.csv")ANCOVA Acquisition Saccade Amplitude
# transform wide format data into long format for mixed ANCOVA
df_long_acq_sacc_amplitude <- melt(df, id = c("id", "iu_group", "sticsa_total"),
measure.vars = c("acq_csp_sacc_amplitude",
"acq_csm_sacc_amplitude"))
# rename columns for easier interpretation
colnames(df_long_acq_sacc_amplitude) = c("id", "iu_group", "sticsa_total", "condition", "sacc_amplitude")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_acq_sacc_amplitude$stimulus <-
factor(ifelse(df_long_acq_sacc_amplitude$condition == "acq_csp_sacc_amplitude", 1, -1))
# mean centre continuous covariate (STICSA)
# to apply mean centring, first obtain average sticsa scores for all participants,
# and save as a variable
df_long_acq_sacc_amplitude$sticsa_total_avg <- mean(df_long_acq_sacc_amplitude$sticsa_total)
# next, subtract this average from all participants' sticsa scores,
# and save as a variable
df_long_acq_sacc_amplitude$sticsa_total_centred <-
df_long_acq_sacc_amplitude$sticsa_total - df_long_acq_sacc_amplitude$sticsa_total_avg
# from this we have mean sticsa scores after centring
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) mixed ANCOVA,
# with mean-cenred STICSA as covariate
# and obtain effect size (partial eta squared)
acq_sacc_amplitude_ancova <-
anova_test(df_long_acq_sacc_amplitude, sacc_amplitude ~ iu_group * stimulus + Error(id/stimulus),
covariate = sticsa_total_centred, effect.size = "pes") ## Warning: NA detected in rows: 234,259.
## Removing this rows before the analysis.# obtain the mixed ANCOVA results
get_anova_table(acq_sacc_amplitude_ancova)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 sticsa_total_centred 1 134 0.007 0.935 0.0000503
## 2 iu_group 1 134 2.128 0.147 0.0160000
## 3 stimulus 1 134 0.943 0.333 0.0070000
## 4 sticsa_total_centred:stimulus 1 134 0.643 0.424 0.0050000
## 5 iu_group:stimulus 1 134 0.864 0.354 0.0060000# results:
# STICSA (centred): F(1,134) = 0.01, p = .935, eta2(partial) < .001
# IU: F(1,134) = 2.13, p = .147, eta2(partial) = .016
# Stimulus: F(1,134) = 0.94, p = .333, eta2(partial) = .007
# STICSA * Stimulus: F(1,134) = 0.64, p = .424, eta2(partial) = .005
# IU * Stimulus: F(1, 134) = 0.86, p = .354, eta2(partial) = .006
# therefore, after accounting for trait anxiety, there continue not
# to be any significant effects of IU, stimulus, and interaction
# effects on saccade amplitude in acquisition.
# write to csv
write.csv((get_anova_table(acq_sacc_amplitude_ancova)),
file = "tables/ancovas/acq_sacc_amplitude_ancova.csv")ANCOVA Extinction Fixation Count
# transform wide format data into long format for mixed ANOVA
df_long_ext_fix_count <- melt(df, id = c("id", "iu_group", "sticsa_total"),
measure.vars = c("e_ext_csp_fix_count",
"e_ext_csm_fix_count",
"l_ext_csp_fix_count",
"l_ext_csm_fix_count"))
# rename columns for easier interpretation
colnames(df_long_ext_fix_count) = c("id", "iu_group", "sticsa_total", "condition", "fix_count")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_ext_fix_count$stimulus <-
factor(ifelse(df_long_ext_fix_count$condition == "e_ext_csp_fix_count" |
df_long_ext_fix_count$condition == "l_ext_csp_fix_count", 1, -1))
# create column to code extinction as early (1) and late (-1)
df_long_ext_fix_count$time <-
factor(ifelse(df_long_ext_fix_count$condition == "e_ext_csp_fix_count" |
df_long_ext_fix_count$condition == "e_ext_csm_fix_count", 1, -1))
# mean centre continuous covariate (STICSA)
# to apply mean centring, first obtain average sticsa scores for all participants,
# and save as a variable
df_long_ext_fix_count$sticsa_total_avg <- mean(df_long_ext_fix_count$sticsa_total)
# next, subtract this average from all participants' sticsa scores,
# and save as a variable
df_long_ext_fix_count$sticsa_total_centred <-
df_long_ext_fix_count$sticsa_total - df_long_ext_fix_count$sticsa_total_avg
# from this we have mean sticsa scores after centring
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) x 2 (Time: Early, Late) mixed ANOVA,
# with mean-centred STICSA as covariate,
# and obtain effect size (partial eta squared)
ext_fix_count_ancova <-
anova_test(df_long_ext_fix_count,
fix_count ~ iu_group * stimulus * time + Error(id/(stimulus*time)),
covariate = sticsa_total_centred, effect.size = "pes")
# obtain the mixed ANCOVA results
get_anova_table(ext_fix_count_ancova)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 sticsa_total_centred 1 136 0.433000 0.512 0.00300000
## 2 iu_group 1 136 4.361000 0.039 * 0.03100000
## 3 stimulus 1 136 4.209000 0.042 * 0.03000000
## 4 time 1 136 5.692000 0.018 * 0.04000000
## 5 sticsa_total_centred:stimulus 1 136 1.098000 0.297 0.00800000
## 6 iu_group:stimulus 1 136 4.560000 0.035 * 0.03200000
## 7 sticsa_total_centred:time 1 136 0.000429 0.984 0.00000316
## 8 iu_group:time 1 136 3.489000 0.064 0.02500000
## 9 stimulus:time 1 136 0.066000 0.797 0.00048800
## 10 sticsa_total_centred:stimulus:time 1 136 0.901000 0.344 0.00700000
## 11 iu_group:stimulus:time 1 136 0.044000 0.834 0.00032500# results:
# STICSA (centred): F(1,136) = 0.43, p = .512, eta2(partial) = .003
# IU: F(1,136) = 4.36, p = .039*, eta2(partial) = .031
# Stimulus: F(1,136) = 4.21, p = .042*, eta2(partial) = .030
# Time: F(1,136) = 5.69, p = .018 *, eta2(partial) = .040
# STICSA * Stimulus: F(1,136) = 1.10, p = .297, eta2(partial) = .008
# IU * Stimulus: F(1, 136) = 4.56, p = .035*, eta2(partial) = .032
# STICSA* Time: F(1,136) = 0.00, p = .982, eta2(partial) < .001
# IU * Time: F(1,136) = 3.49, p = .064, eta2(partial) = . 025
# Stimulus * Time: F(1,136) = 0.07, p = .797, eta2(partial) < .001
# STICSA * Stimulus * Time: F(1,136) = 0.90, p = .344, eta2(partial) = .007
# IU * Stimulus * Time: F(1,136) = 0.04, p = .834, eta2(partial) < .001
# therefore, after accounting for trait anxiety, IU, Stimulus, and Time
# continue to have a significant effect on fixation duration in acquisition.
# there is no longer a significant interaction effect of IU*Time,
# but there is now a significant interaction effect of IU*stimulus
# write to csv
write.csv((get_anova_table(ext_fix_count_ancova)),
file = "tables/ancovas/ext_fix_count_ancova.csv")
# as there was a significant IU*Stimulus interaction, conduct simple
# main effects analysis:
## obtain effect of IU at each level of stimulus
simple_effects_ext_fix_count_iu_ancova <- df_long_ext_fix_count %>%
group_by(stimulus) %>%
anova_test(dv = fix_count, wid = id, between = iu_group, within = time,
covariate = sticsa_total_centred, effect.size = "pes") %>%
get_anova_table()
# get the output
simple_effects_ext_fix_count_iu_ancova## # A tibble: 10 × 8
## stimulus Effect DFn DFd F p `p<.05` pes
## * <fct> <chr> <dbl> <dbl> <dbl> <dbl> <chr> <dbl>
## 1 -1 sticsa_total_centred 1 136 0.142 0.707 "" 0.001
## 2 -1 iu_group 1 136 6.66 0.011 "*" 0.047
## 3 -1 time 1 136 5.02 0.027 "*" 0.036
## 4 -1 sticsa_total_centred:time 1 136 0.369 0.545 "" 0.003
## 5 -1 iu_group:time 1 136 2.25 0.136 "" 0.016
## 6 1 sticsa_total_centred 1 136 0.796 0.374 "" 0.006
## 7 1 iu_group 1 136 2.16 0.143 "" 0.016
## 8 1 time 1 136 2.86 0.093 "" 0.021
## 9 1 sticsa_total_centred:time 1 136 0.253 0.616 "" 0.002
## 10 1 iu_group:time 1 136 2.40 0.124 "" 0.017# results:
# The effect of IU group on CS+ was not significant [F(1,136) = 2.17, p = .143, pes = .016]
# the effect of IU group on CS- was significant [F(1,136) = 6.66, p = .011, pes = .047]
# as there were a significant main effects of IU, stimulus and time,
# as well as a sig IU-STimulus interaction, obtain estimated
# marginal means to be reported (temporarily from SPSS):
# IU
emmeans_ext_fix_count_ancova_high_iu <- 8.15
emmeans_ext_fix_count_ancova_low_iu <- 6.87
# stimulus
emmeans_ext_fix_count_ancova_csp <- 7.37
emmeans_ext_fix_count_ancova_csm <- 7.65
# time
emmeans_ext_fix_count_ancova_early <- 7.29
emmeans_ext_fix_count_ancova_late <- 7.72
# IU-Stimulus interaction
emmeans_ext_fix_count_ancova_high_iu_csp <- 7.83
emmeans_ext_fix_count_ancova_high_iu_csm <- 8.46
emmeans_ext_fix_count_ancova_low_iu_csp <- 6.90
emmeans_ext_fix_count_ancova_low_iu_csm <- 6.84ANCOVA Extinction Fixation Duration (Log Transformed)
# transform wide format data into long format for mixed ANCOVA
df_long_ext_fix_duration_log <- melt(df, id = c("id", "iu_group", "sticsa_total"),
measure.vars = c("e_ext_csp_fix_duration_log",
"e_ext_csm_fix_duration_log",
"l_ext_csp_fix_duration_log",
"l_ext_csm_fix_duration_log"))
# rename columns for easier interpretation
colnames(df_long_ext_fix_duration_log) = c("id", "iu_group", "sticsa_total", "condition", "fix_duration_log")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_ext_fix_duration_log$stimulus <-
factor(ifelse(df_long_ext_fix_duration_log$condition == "e_ext_csp_fix_duration_log" |
df_long_ext_fix_duration_log$condition == "l_ext_csp_fix_duration_log", 1, -1))
# create column to code extinction as early (1) and late (-1)
df_long_ext_fix_duration_log$time <-
factor(ifelse(df_long_ext_fix_duration_log$condition == "e_ext_csp_fix_duration_log" |
df_long_ext_fix_duration_log$condition == "e_ext_csm_fix_duration_log", 1, -1))
# mean centre continuous covariate (STICSA)
# to apply mean centring, first obtain average sticsa scores for all participants,
# and save as a variable
df_long_ext_fix_duration_log$sticsa_total_avg <- mean(df_long_ext_fix_duration_log$sticsa_total)
# next, subtract this average from all participants' sticsa scores,
# and save as a variable
df_long_ext_fix_duration_log$sticsa_total_centred <-
df_long_ext_fix_duration_log$sticsa_total - df_long_ext_fix_duration_log$sticsa_total_avg
# from this we have mean sticsa scores after centring
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) x 2 (Time: Early, Late) mixed ANOVA,
# with mean-centred STICSA as covariate,
# and obtain effect size (partial eta squared)
ext_fix_duration_ancova_log <-
anova_test(df_long_ext_fix_duration_log,
fix_duration_log ~ iu_group * stimulus * time + Error(id/(stimulus*time)),
covariate = sticsa_total_centred, effect.size = "pes")
# obtain the mixed ANCOVA results
get_anova_table(ext_fix_duration_ancova_log)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 sticsa_total_centred 1 136 0.001 0.972 0.00000901
## 2 iu_group 1 136 8.365 0.004 * 0.05800000
## 3 stimulus 1 136 0.514 0.475 0.00400000
## 4 time 1 136 4.358 0.039 * 0.03100000
## 5 sticsa_total_centred:stimulus 1 136 0.195 0.659 0.00100000
## 6 iu_group:stimulus 1 136 5.357 0.022 * 0.03800000
## 7 sticsa_total_centred:time 1 136 0.329 0.567 0.00200000
## 8 iu_group:time 1 136 0.501 0.480 0.00400000
## 9 stimulus:time 1 136 0.174 0.677 0.00100000
## 10 sticsa_total_centred:stimulus:time 1 136 0.221 0.639 0.00200000
## 11 iu_group:stimulus:time 1 136 0.379 0.539 0.00300000# results:
# STICSA (centred): F(1,136) = 0.01, p = .972, eta2(partial) < .001
# IU: F(1,136) = 8.37, p = .004**, eta2(partial) = .058
# Stimulus: F(1,136) = 0.51, p = .475, eta2(partial) = .004
# Time: F(1,136) = 4.36, p = .039*, eta2(partial) = .031
# STICSA * Stimulus: F(1,136) = 0.20, p = .659, eta2(partial) = .001
# IU * Stimulus: F(1, 136) = 5.36, p = .022*, eta2(partial) = .038
# STICSA* Time: F(1,136) = 0.33, p = .567, eta2(partial) = .002
# IU * Time: F(1,136) = 0.50, p = .480, eta2(partial) = . 004
# Stimulus * Time: F(1,136) = 0.17, p = .677, eta2(partial) = .001
# STICSA * Stimulus * Time: F(1,136) = 0.22, p = .639, eta2(partial) = .002
# IU * Stimulus * Time: F(1,136) = 0.34, p = .539, eta2(partial) = .003
# there were significant main effects of IU, time,
# and a significant IU-stimulus interaction on fixation duration in extinction,
# and no further main effects or interactions.
# write to csv
write.csv((get_anova_table(ext_fix_duration_ancova_log)),
file = "tables/ancovas/ext_fix_duration_ancova_log.csv")
# as there was a significant IU*Stimulus interaction, conduct simple
# main effects analysis:
## obtain effect of IU at each level of stimulus
simple_effects_ext_fix_duration_iu_ancova <- df_long_ext_fix_duration_log %>%
group_by(stimulus) %>%
anova_test(dv = fix_duration_log, wid = id, between = iu_group, within = time,
covariate = sticsa_total_centred, effect.size = "pes") %>%
get_anova_table()
# get the output
simple_effects_ext_fix_duration_iu_ancova## # A tibble: 10 × 8
## stimulus Effect DFn DFd F p `p<.05` pes
## * <fct> <chr> <dbl> <dbl> <dbl> <dbl> <chr> <dbl>
## 1 -1 sticsa_total_centred 1 136 0.008 0.928 "" 0.0000602
## 2 -1 iu_group 1 136 11.2 0.001 "*" 0.076
## 3 -1 time 1 136 4.37 0.038 "*" 0.031
## 4 -1 sticsa_total_centred:time 1 136 0.627 0.43 "" 0.005
## 5 -1 iu_group:time 1 136 0.061 0.806 "" 0.000445
## 6 1 sticsa_total_centred 1 136 0.027 0.87 "" 0.000199
## 7 1 iu_group 1 136 4.70 0.032 "*" 0.033
## 8 1 time 1 136 1.93 0.167 "" 0.014
## 9 1 sticsa_total_centred:time 1 136 0.036 0.849 "" 0.000266
## 10 1 iu_group:time 1 136 0.771 0.381 "" 0.006# results:
# The effect of IU group on CS+ was not significant [F(1,136) = 4.70, p = .032, pes = .033]
# the effect of IU group on CS- was significant [F(1,136) = 11,19, p = .001, pes = .076]
# as there were a significant main effects of IU, time and iu-stimulus interaction,
# obtain estimated marginal means to be reported (temporarily from SPSS):
##### NOTE: have not done log-transformed fixation duration on SPSS yet, and
# so do not have estimated marginal means for fixation duration (log-transformed)
# IU
emmeans_ext_fix_duration_ancova_high_iu <- 0
emmeans_ext_fix_duration_ancova_low_iu <- 0
# time
emmeans_ext_fix_duration_ancova_early <- 0
emmeans_ext_fix_duration_ancova_late <- 0
# IU-Stimulus interaction
emmeans_ext_fix_duration_ancova_high_iu_csp <- 0
emmeans_ext_fix_duration_ancova_high_iu_csm <- 0
emmeans_ext_fix_duration_ancova_low_iu_csp <- 0
emmeans_ext_fix_duration_ancova_low_iu_csm <- 0ANCOVA Extinction Saccade Amplitude
# transform wide format data into long format for mixed ANCOVA
df_long_ext_sacc_amplitude <- melt(df, id = c("id", "iu_group", "sticsa_total"),
measure.vars = c("e_ext_csp_sacc_amplitude",
"e_ext_csm_sacc_amplitude",
"l_ext_csp_sacc_amplitude",
"l_ext_csm_sacc_amplitude"))
# rename columns for easier interpretation
colnames(df_long_ext_sacc_amplitude) = c("id", "iu_group", "sticsa_total", "condition", "sacc_amplitude")
# create column to code stimulus as CS+ (1) and CS- (-1)
df_long_ext_sacc_amplitude$stimulus <-
factor(ifelse(df_long_ext_sacc_amplitude$condition == "e_ext_csp_sacc_amplitude" |
df_long_ext_sacc_amplitude$condition == "l_ext_csp_sacc_amplitude", 1, -1))
# create column to code extinction as early (1) and late (-1)
df_long_ext_sacc_amplitude$time <-
factor(ifelse(df_long_ext_sacc_amplitude$condition == "e_ext_csp_sacc_amplitude" |
df_long_ext_sacc_amplitude$condition == "e_ext_csm_sacc_amplitude", 1, -1))
# mean centre continuous covariate (STICSA)
# to apply mean centring, first obtain average sticsa scores for all participants,
# and save as a variable
df_long_ext_sacc_amplitude$sticsa_total_avg <- mean(df_long_ext_sacc_amplitude$sticsa_total)
# next, subtract this average from all participants' sticsa scores,
# and save as a variable
df_long_ext_sacc_amplitude$sticsa_total_centred <-
df_long_ext_sacc_amplitude$sticsa_total - df_long_ext_sacc_amplitude$sticsa_total_avg
# from this we have mean sticsa scores after centring
# compute 2(IU: High & Low) x 2 (Stimulus: CS+, CS-) x 2 (Time: Early, Late) mixed ANOVA,
# with mean-centred STICSA as covariate,
# and obtain effect size (partial eta squared)
ext_sacc_amplitude_ancova <-
anova_test(df_long_ext_sacc_amplitude,
sacc_amplitude ~ iu_group * stimulus * time + Error(id/(stimulus*time)),
covariate = sticsa_total_centred, effect.size = "pes") ## Warning: NA detected in rows: 116,181,301.
## Removing this rows before the analysis.# obtain the mixed ANCOVA results
get_anova_table(ext_sacc_amplitude_ancova)## ANOVA Table (type III tests)
##
## Effect DFn DFd F p p<.05 pes
## 1 sticsa_total_centred 1 133 1.134 0.289 0.008000
## 2 iu_group 1 133 1.025 0.313 0.008000
## 3 stimulus 1 133 0.754 0.387 0.006000
## 4 time 1 133 0.255 0.615 0.002000
## 5 sticsa_total_centred:stimulus 1 133 0.370 0.544 0.003000
## 6 iu_group:stimulus 1 133 2.035 0.156 0.015000
## 7 sticsa_total_centred:time 1 133 1.359 0.246 0.010000
## 8 iu_group:time 1 133 0.803 0.372 0.006000
## 9 stimulus:time 1 133 0.071 0.790 0.000533
## 10 sticsa_total_centred:stimulus:time 1 133 0.421 0.517 0.003000
## 11 iu_group:stimulus:time 1 133 0.997 0.320 0.007000# results:
# STICSA (centred): F(1,133) = 1.13, p = .289, eta2(partial) = .008
# IU: F(1,133) = 1.03, p = .313, eta2(partial) = .008
# Stimulus: F(1,133) = 0.75, p = .387, eta2(partial) = .006
# Time: F(1,133) = 0.26, p = .615, eta2(partial) = .002
# STICSA * Stimulus: F(1,133) = 0.37, p = .544, eta2(partial) = .003
# IU * Stimulus: F(1, 133) = 2.04, p = .156, eta2(partial) = .015
# STICSA* Time: F(1,133) = 1.36, p = .246, eta2(partial) = .010
# IU * Time: F(1,133) = 0.80, p = .372, eta2(partial) = . 006
# Stimulus * Time: F(1,133) = 0.07, p = .790, eta2(partial) = .001
# STICSA * Stimulus * Time: F(1,133) = 0.42, p = .517, eta2(partial) = .003
# IU * Stimulus * Time: F(1,133) = 0.10, p = .320, eta2(partial) = .007
# therefore, even after accounting for trait anxiety, there continue
# to be no significant effects or interactions on saccade amplitude
# in extinction
# write to csv
write.csv((get_anova_table(ext_sacc_amplitude_ancova)),
file = "tables/ancovas/ext_sacc_amplitude_ancova.csv")Assumption Checks
# assumptions of mixed ANOVA:
## categorical IVs, interval/ratio DVs
## residuals approximate normal distribution (for each level of each IV)
## homogeneity of variances
## homoscedasticity (plot standardised residuals against predicted values)
## sphericity (not applicable in this case, as no within-subjects factors with > 3 levels)
## homogeneity of variance-covariance matrices (this is assumed as sample size for
##### each group and testing session is roughly equal)
# additional assumptions of ANCOVA:
## independence of covariate and IVs
## homogeneity of regression slopes
## linearity between covariate and outcome variable(s) at each level of grouping
#### variable(s). check by creating grouped scatterplot
## outcome variable(s) should be approximately normally distributed
## no sigifnicant outliers in the groupsLinearity Between Covariate and Outcome Variables
## this is at each level of grouping variable.
# check by computing grouped scatterplot of covariate and outcome variable
################# linearity
# linearity between covariate and outcome variable at each level of grouping variable
# (check using grouped scatterplot of covariate and outcome variable)
### iu group
scatterplot_acq_fix_count_sticsa_centred_by_iu <- ggplot(df_long_acq_fix_count,
aes(x = sticsa_total_centred, y = fix_count,
colour = iu_group)) +
geom_point() +
geom_jitter(width = .5, alpha = .30, size = 2.5) +
geom_smooth(method = lm, se = FALSE) +
labs(title = "Plot of the Relationship Between Trait Anxiety (Covariate) and
Fixation Count in Acquisition by IU Group",
x = "STICSA Total (Centred)",
y = "Acquisition Fixation Count") +
theme_classic() +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
theme(text = element_text(family = "serif")) +
guides(colour = guide_legend(reverse = TRUE)) +
scale_colour_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(colour = "IU Group") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold"))
print(scatterplot_acq_fix_count_sticsa_centred_by_iu)## `geom_smooth()` using formula 'y ~ x'
# there appears to be a linear relationship between the covariate (sticsa total centred)
# and outcome variable (fixation count acquisition) at both levels of the grouping variable (iu group)
# however, this plot also demonstrates likely heterogeneity of regression slopes, as there is possibly an
# interaction between the outcome and covariate, and the regression lines do not appear
# parallel
####### extinctino iu
### iu group
scatterplot_ext_fix_count_sticsa_centred_by_iu <- ggplot(df_long_ext_fix_count,
aes(x = sticsa_total_centred, y = fix_count,
colour = iu_group)) +
geom_point() +
geom_jitter(width = .5, alpha = .30, size = 2.5) +
geom_smooth(method = lm, se = FALSE) +
labs(title = "Plot of the Relationship Between Trait Anxiety (Covariate) and
Fixation Count in Extinction by IU Group and Time",
x = "STICSA Total (Centred)",
y = "Extinction Fixation Count") +
theme_classic() +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
theme(text = element_text(family = "serif")) +
guides(colour = guide_legend(reverse = TRUE)) +
scale_colour_manual(values = c("#c45150", "#824372"), labels = c("Low IU", "High IU")) +
labs(colour = "IU Group") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold")) +
facet_wrap(~ time, labeller = "label_both")
print(scatterplot_ext_fix_count_sticsa_centred_by_iu)## `geom_smooth()` using formula 'y ~ x'
### stimulus
scatterplot_acq_fix_count_sticsa_centred_by_stimulus <- ggplot(df_long_acq_fix_count,
aes(x = sticsa_total_centred, y = fix_count,
colour = stimulus)) +
geom_point() +
geom_jitter(width = .5, alpha = .30, size = 2.5) +
geom_smooth(method = lm, se = FALSE) +
labs(title = "Plot of the Relationship Between Trait Anxiety (Covariate) and
Fixation Count in Extinction by Stimulus",
x = "STICSA Total (Centred)",
y = "Acquisition Fixation Count") +
theme_classic() +
theme(plot.title = element_text(face = "bold", hjust = 0.5, size = 12)) +
theme(text = element_text(family = "serif")) +
guides(colour = guide_legend(reverse = TRUE)) +
scale_colour_manual(values = c("#DCECC9", "#076B68"), labels = c("CS-", "CS+")) +
labs(colour = "Stimulus") +
theme(legend.position = "bottom", legend.title = element_text(face = "bold"))
print(scatterplot_acq_fix_count_sticsa_centred_by_stimulus)## `geom_smooth()` using formula 'y ~ x'
# there appears to be a linear relationship between the covariate (sticsa total centred)
# and outcome variable (fixation count acquisition) at both levels of the grouping variable (stimulus)
# this plot also demonstrates homogeneity of regression slopes, as there is no
# interaction between the outcome and covariate, and the regression lines appear roughly
# parallel
# this also checks homogeneity of regression slopes:
# slopes of regression lines should be same for each group (no interaction between
# covariate and outcome) i.e. the lines should be parallel
## reporting example: There was a linear relationship between pre-test and post-test anxiety score for each training group, as assessed by visual inspection of a scatter plot.Normality
## acquisition fix count
df_long_acq_fix_count %>%
group_by(iu_group, stimulus) %>%
shapiro_test(fix_count)## # A tibble: 4 × 5
## iu_group stimulus variable statistic p
## <fct> <fct> <chr> <dbl> <dbl>
## 1 -1 -1 fix_count 0.929 0.000611
## 2 -1 1 fix_count 0.934 0.00108
## 3 1 -1 fix_count 0.962 0.0364
## 4 1 1 fix_count 0.895 0.0000312## extinction fix count
df_long_ext_fix_count %>%
group_by(iu_group, stimulus, time) %>%
shapiro_test(fix_count)## # A tibble: 8 × 6
## iu_group stimulus time variable statistic p
## <fct> <fct> <fct> <chr> <dbl> <dbl>
## 1 -1 -1 -1 fix_count 0.961 0.0263
## 2 -1 -1 1 fix_count 0.904 0.0000488
## 3 -1 1 -1 fix_count 0.977 0.228
## 4 -1 1 1 fix_count 0.881 0.00000681
## 5 1 -1 -1 fix_count 0.929 0.000810
## 6 1 -1 1 fix_count 0.981 0.391
## 7 1 1 -1 fix_count 0.945 0.00457
## 8 1 1 1 fix_count 0.931 0.000995# acquisition fix duration log
df_long_acq_fix_duration_log %>%
group_by(iu_group, stimulus) %>%
shapiro_test(fix_duration_log)## # A tibble: 4 × 5
## iu_group stimulus variable statistic p
## <fct> <fct> <chr> <dbl> <dbl>
## 1 -1 -1 fix_duration_log 0.970 0.0814
## 2 -1 1 fix_duration_log 0.964 0.0398
## 3 1 -1 fix_duration_log 0.981 0.385
## 4 1 1 fix_duration_log 0.981 0.408## extinction fix count
df_long_ext_fix_duration_log %>%
group_by(iu_group, stimulus, time) %>%
shapiro_test(fix_duration_log)## # A tibble: 8 × 6
## iu_group stimulus time variable statistic p
## <fct> <fct> <fct> <chr> <dbl> <dbl>
## 1 -1 -1 -1 fix_duration_log 0.945 0.00364
## 2 -1 -1 1 fix_duration_log 0.974 0.143
## 3 -1 1 -1 fix_duration_log 0.972 0.112
## 4 -1 1 1 fix_duration_log 0.970 0.0913
## 5 1 -1 -1 fix_duration_log 0.973 0.146
## 6 1 -1 1 fix_duration_log 0.959 0.0242
## 7 1 1 -1 fix_duration_log 0.984 0.523
## 8 1 1 1 fix_duration_log 0.983 0.460## acquisition sacc amplitude
df_long_acq_sacc_amplitude %>%
group_by(iu_group, stimulus) %>%
shapiro_test(sacc_amplitude)## # A tibble: 4 × 5
## iu_group stimulus variable statistic p
## <fct> <fct> <chr> <dbl> <dbl>
## 1 -1 -1 sacc_amplitude 0.940 0.00227
## 2 -1 1 sacc_amplitude 0.954 0.0111
## 3 1 -1 sacc_amplitude 0.913 0.000176
## 4 1 1 sacc_amplitude 0.918 0.000275## extinction sacc amplitude
df_long_ext_sacc_amplitude %>%
group_by(iu_group, stimulus, time) %>%
shapiro_test(sacc_amplitude)## # A tibble: 8 × 6
## iu_group stimulus time variable statistic p
## <fct> <fct> <fct> <chr> <dbl> <dbl>
## 1 -1 -1 -1 sacc_amplitude 0.849 0.000000535
## 2 -1 -1 1 sacc_amplitude 0.889 0.0000125
## 3 -1 1 -1 sacc_amplitude 0.821 0.0000000925
## 4 -1 1 1 sacc_amplitude 0.880 0.00000688
## 5 1 -1 -1 sacc_amplitude 0.930 0.000946
## 6 1 -1 1 sacc_amplitude 0.902 0.0000659
## 7 1 1 -1 sacc_amplitude 0.925 0.000514
## 8 1 1 1 sacc_amplitude 0.926 0.000578## however, as sample size is greater than 50, normal QQ plot is preferred because
# at larger sample sizes, Shapiro-Wilk becomes very sensitive to even a minor
# deviation from normality.
# (q-q plot draw correlation btw data and the normal distribution)
ggqqplot(df_long_acq_fix_count, "fix_count", ggtheme = theme_classic()) +
facet_grid(stimulus ~ iu_group, labeller = "label_both")
ggqqplot(df_long_ext_fix_count, "fix_count", ggtheme = theme_classic()) +
facet_grid(stimulus + time ~ iu_group, labeller = "label_both")
### reporting example:
# The score were normally distributed (p > 0.05) for each cell, as assessed by Shapiro-Wilk’s test of normality.
# All the points fall approximately along the reference line, for each cell. So we can assume normality of the data.Normality of Residuals
# first compute model using lm(). covariate goes first.f
ancova_lm_model_acq_fix_count <-
lm(fix_count ~ sticsa_total_centred + iu_group * stimulus, df_long_acq_fix_count)
# augment data by adding fitted values and residuals
model_metrics <- augment(ancova_lm_model_acq_fix_count)
# assess normality of residuals using shapiro wilk
shapiro_test(model_metrics$.resid) # not normally distributed## # A tibble: 1 × 3
## variable statistic p.value
## <chr> <dbl> <dbl>
## 1 model_metrics$.resid 0.940 0.00000000340# reporting:
# The Shapiro Wilk test was not significant (p > 0.05), so we can assume normality of residuals
###### also outliers:
# examine standardised/studentised residual. observations whose standardised residuals
# are greater than 3 in absolute value are possible outliers.
model_metrics %>%
filter(abs(.std.resid) > 3) %>%
as.data.frame()## fix_count sticsa_total_centred iu_group stimulus .fitted .resid .hat
## 1 23.16667 27.46043 1 1 7.697435 15.46923 0.04169528
## 2 20.16667 11.46043 1 1 7.564182 12.60248 0.01706000
## 3 18.33333 27.46043 1 -1 8.159690 10.17364 0.04169528
## .sigma .cooksd .std.resid
## 1 3.286205 0.18613043 4.624897
## 2 3.335139 0.04804383 3.720299
## 3 3.364537 0.08050687 3.041654# reporting example: There were no outliers in the data, as assessed by no cases with standardized residuals greater than 3 in absolute value.Homogeneity of Variance
# this will be done using levene's test
# however, in large samples, levene's test can be sig even when group variances
# are not very different.
## acquisition fix count
df_long_acq_fix_count %>%
group_by(stimulus) %>%
levene_test(fix_count ~ iu_group)## # A tibble: 2 × 5
## stimulus df1 df2 statistic p
## <fct> <int> <int> <dbl> <dbl>
## 1 -1 1 137 0.477 0.491
## 2 1 1 137 0.0415 0.839## extinction fix count
df_long_ext_fix_count %>%
group_by(stimulus, time) %>%
levene_test(fix_count ~ iu_group)## # A tibble: 4 × 6
## stimulus time df1 df2 statistic p
## <fct> <fct> <int> <int> <dbl> <dbl>
## 1 -1 -1 1 137 1.45 0.231
## 2 -1 1 1 137 0.181 0.671
## 3 1 -1 1 137 0.264 0.608
## 4 1 1 1 137 1.86 0.174# acquisition fix duration log
df_long_acq_fix_duration_log %>%
group_by(stimulus) %>%
levene_test(fix_duration_log ~ iu_group)## # A tibble: 2 × 5
## stimulus df1 df2 statistic p
## <fct> <int> <int> <dbl> <dbl>
## 1 -1 1 137 2.04 0.155
## 2 1 1 137 0.753 0.387## extinction fix count
df_long_ext_fix_duration_log %>%
group_by(stimulus, time) %>%
levene_test(fix_duration_log ~ iu_group)## # A tibble: 4 × 6
## stimulus time df1 df2 statistic p
## <fct> <fct> <int> <int> <dbl> <dbl>
## 1 -1 -1 1 137 8.18 0.00490
## 2 -1 1 1 137 7.74 0.00616
## 3 1 -1 1 137 2.78 0.0977
## 4 1 1 1 137 7.14 0.00843## acquisition sacc amplitude
df_long_acq_sacc_amplitude %>%
group_by(stimulus) %>%
levene_test(sacc_amplitude ~ iu_group)## # A tibble: 2 × 5
## stimulus df1 df2 statistic p
## <fct> <int> <int> <dbl> <dbl>
## 1 -1 1 135 1.03 0.311
## 2 1 1 137 3.42 0.0665## extinction sacc amplitude
df_long_ext_sacc_amplitude %>%
group_by(stimulus, time) %>%
levene_test(sacc_amplitude ~ iu_group)## # A tibble: 4 × 6
## stimulus time df1 df2 statistic p
## <fct> <fct> <int> <int> <dbl> <dbl>
## 1 -1 -1 1 137 0.364 0.547
## 2 -1 1 1 136 1.72 0.191
## 3 1 -1 1 136 0.0230 0.880
## 4 1 1 1 136 0.0324 0.857### reproting: For IUS total, the variances were not similar for the
# High and Low IU groups, F(1,137) = 6.75, p = .010
# (i.e. spread of IU scores is different between groups)
# There was homogeneity of variances, as assessed by Levene’s test (p > 0.05).Homoscedasticity
# aka homogeneity of residuals variance for all groups. the residuals are
# assumed to have a constant variance (homoscedasticity)Homogeneity of Variance-Covariance Matrices
# this tests whether covariance matrices are equal across cells formed by
# between-subjects factor (IU)
# use Box's M (however, this is highly sensitive, so unless p < .001 and sample
# sizes are unequal, can ignore it)
box_m(df_long_acq_fix_count[, "fix_count", drop = FALSE], df_long_acq_fix_count$iu_group)## # A tibble: 1 × 4
## statistic p.value parameter method
## <dbl> <dbl> <dbl> <chr>
## 1 0.224 0.636 1 Box's M-test for Homogeneity of Covariance Matric…box_m(df_long_ext_fix_count[, "fix_count", drop = FALSE], df_long_ext_fix_count$iu_group)## # A tibble: 1 × 4
## statistic p.value parameter method
## <dbl> <dbl> <dbl> <chr>
## 1 0.753 0.385 1 Box's M-test for Homogeneity of Covariance Matric…box_m(df_long_acq_fix_duration_log[, "fix_duration_log", drop = FALSE], df_long_acq_fix_duration_log$iu_group)## # A tibble: 1 × 4
## statistic p.value parameter method
## <dbl> <dbl> <dbl> <chr>
## 1 0.358 0.550 1 Box's M-test for Homogeneity of Covariance Matric…box_m(df_long_ext_fix_duration_log[, "fix_duration_log", drop = FALSE], df_long_ext_fix_duration_log$iu_group)## # A tibble: 1 × 4
## statistic p.value parameter method
## <dbl> <dbl> <dbl> <chr>
## 1 16.7 0.0000435 1 Box's M-test for Homogeneity of Covariance Matr…box_m(df_long_acq_sacc_amplitude[, "sacc_amplitude", drop = FALSE], df_long_acq_sacc_amplitude$iu_group)## # A tibble: 1 × 4
## statistic p.value parameter method
## <dbl> <dbl> <dbl> <chr>
## 1 NA NA 1 Box's M-test for Homogeneity of Covariance Matric…box_m(df_long_ext_sacc_amplitude[, "sacc_amplitude", drop = FALSE], df_long_ext_sacc_amplitude$iu_group)## # A tibble: 1 × 4
## statistic p.value parameter method
## <dbl> <dbl> <dbl> <chr>
## 1 NA NA 1 Box's M-test for Homogeneity of Covariance Matric…## reporting example: There was homogeneity of covariances, as assessed by Box’s test of equality of covariance matrices (p > 0.001).Independence of Covariate and IVs
Fixation Count
Acquisition
# sticsa and iu group
t_test_independence_sticsa_iu_group_acq_fix_count <-
t.test(
df_long_acq_fix_count[df_long_acq_fix_count$iu_group == "1", "sticsa_total_centred"],
df_long_acq_fix_count[df_long_acq_fix_count$iu_group == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_iu_group_acq_fix_count##
## Two Sample t-test
##
## data: df_long_acq_fix_count[df_long_acq_fix_count$iu_group == "1", "sticsa_total_centred"] and df_long_acq_fix_count[df_long_acq_fix_count$iu_group == "-1", "sticsa_total_centred"]
## t = 9.3255, df = 276, p-value < 0.00000000000000022
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 7.343247 11.273157
## sample estimates:
## mean of x mean of y
## 4.754549 -4.553653# p < .05 : sticsa is not independent of iu group
# sticsa and stimulus
t_test_independence_sticsa_stimulus_acq_fix_count <-
t.test(
df_long_acq_fix_count[df_long_acq_fix_count$stimulus == "1", "sticsa_total_centred"],
df_long_acq_fix_count[df_long_acq_fix_count$stimulus == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_stimulus_acq_fix_count##
## Two Sample t-test
##
## data: df_long_acq_fix_count[df_long_acq_fix_count$stimulus == "1", "sticsa_total_centred"] and df_long_acq_fix_count[df_long_acq_fix_count$stimulus == "-1", "sticsa_total_centred"]
## t = 0, df = 276, p-value = 1
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -2.252832 2.252832
## sample estimates:
## mean of x mean of y
## -0.000000000000002862893 -0.000000000000002862893# p > .05 - sticsa is independent of stimulusExtinction
# sticsa and iu group
t_test_independence_sticsa_iu_group_ext_fix_count <-
t.test(
df_long_ext_fix_count[df_long_ext_fix_count$iu_group == "1", "sticsa_total_centred"],
df_long_ext_fix_count[df_long_ext_fix_count$iu_group == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_iu_group_ext_fix_count##
## Two Sample t-test
##
## data: df_long_ext_fix_count[df_long_ext_fix_count$iu_group == "1", "sticsa_total_centred"] and df_long_ext_fix_count[df_long_ext_fix_count$iu_group == "-1", "sticsa_total_centred"]
## t = 13.212, df = 554, p-value < 0.00000000000000022
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 7.924338 10.692067
## sample estimates:
## mean of x mean of y
## 4.754549 -4.553653# p < .05 : sticsa is not independent of iu group
# sticsa and stimulus
t_test_independence_sticsa_stimulus_ext_fix_count <-
t.test(
df_long_ext_fix_count[df_long_ext_fix_count$stimulus == "1", "sticsa_total_centred"],
df_long_ext_fix_count[df_long_ext_fix_count$stimulus == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_stimulus_ext_fix_count##
## Two Sample t-test
##
## data: df_long_ext_fix_count[df_long_ext_fix_count$stimulus == "1", "sticsa_total_centred"] and df_long_ext_fix_count[df_long_ext_fix_count$stimulus == "-1", "sticsa_total_centred"]
## t = 0, df = 554, p-value = 1
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -1.586608 1.586608
## sample estimates:
## mean of x mean of y
## -0.000000000000002862855 -0.000000000000002862855# p > .05 - sticsa is independent of stimulus
# sticsa and time
t_test_independence_sticsa_time_ext_fix_count <-
t.test(
df_long_ext_fix_count[df_long_ext_fix_count$time == "1", "sticsa_total_centred"],
df_long_ext_fix_count[df_long_ext_fix_count$time == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_time_ext_fix_count##
## Two Sample t-test
##
## data: df_long_ext_fix_count[df_long_ext_fix_count$time == "1", "sticsa_total_centred"] and df_long_ext_fix_count[df_long_ext_fix_count$time == "-1", "sticsa_total_centred"]
## t = 0, df = 554, p-value = 1
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -1.586608 1.586608
## sample estimates:
## mean of x mean of y
## -0.000000000000002862855 -0.000000000000002862855# p > .05 - sticsa is independent of timeFixation Duration (Log Transformed)
Acquisition
# sticsa and iu group
t_test_independence_sticsa_iu_group_acq_fix_duration_log <-
t.test(
df_long_acq_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "1", "sticsa_total_centred"],
df_long_acq_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_iu_group_acq_fix_duration_log##
## Two Sample t-test
##
## data: df_long_acq_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "1", "sticsa_total_centred"] and df_long_acq_fix_duration_log[df_long_acq_fix_duration_log$iu_group == "-1", "sticsa_total_centred"]
## t = 9.3255, df = 276, p-value < 0.00000000000000022
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 7.343247 11.273157
## sample estimates:
## mean of x mean of y
## 4.754549 -4.553653# p < .05 : sticsa is not independent of iu group
# sticsa and stimulus
t_test_independence_sticsa_stimulus_acq_fix_duration_log <-
t.test(
df_long_acq_fix_duration_log[df_long_acq_fix_duration_log$stimulus == "1", "sticsa_total_centred"],
df_long_acq_fix_duration_log[df_long_acq_fix_duration_log$stimulus == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_stimulus_acq_fix_duration_log##
## Two Sample t-test
##
## data: df_long_acq_fix_duration_log[df_long_acq_fix_duration_log$stimulus == "1", "sticsa_total_centred"] and df_long_acq_fix_duration_log[df_long_acq_fix_duration_log$stimulus == "-1", "sticsa_total_centred"]
## t = 0, df = 276, p-value = 1
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -2.252832 2.252832
## sample estimates:
## mean of x mean of y
## -0.000000000000002862893 -0.000000000000002862893# p > .05 - sticsa is independent of stimulusExtinction
# sticsa and iu group
t_test_independence_sticsa_iu_group_ext_fix_duration_log <-
t.test(
df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "1", "sticsa_total_centred"],
df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_iu_group_ext_fix_duration_log##
## Two Sample t-test
##
## data: df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "1", "sticsa_total_centred"] and df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$iu_group == "-1", "sticsa_total_centred"]
## t = 13.212, df = 554, p-value < 0.00000000000000022
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 7.924338 10.692067
## sample estimates:
## mean of x mean of y
## 4.754549 -4.553653# p < .05 : sticsa is not independent of iu group
# sticsa and stimulus
t_test_independence_sticsa_stimulus_ext_fix_duration_log <-
t.test(
df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$stimulus == "1", "sticsa_total_centred"],
df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$stimulus == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_stimulus_ext_fix_duration_log##
## Two Sample t-test
##
## data: df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$stimulus == "1", "sticsa_total_centred"] and df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$stimulus == "-1", "sticsa_total_centred"]
## t = 0, df = 554, p-value = 1
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -1.586608 1.586608
## sample estimates:
## mean of x mean of y
## -0.000000000000002862855 -0.000000000000002862855# p > .05 - sticsa is independent of stimulus
# sticsa and time
t_test_independence_sticsa_time_ext_fix_duration <-
t.test(
df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$time == "1", "sticsa_total_centred"],
df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$time == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_time_ext_fix_duration##
## Two Sample t-test
##
## data: df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$time == "1", "sticsa_total_centred"] and df_long_ext_fix_duration_log[df_long_ext_fix_duration_log$time == "-1", "sticsa_total_centred"]
## t = 0, df = 554, p-value = 1
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -1.586608 1.586608
## sample estimates:
## mean of x mean of y
## -0.000000000000002862855 -0.000000000000002862855# p > .05 - sticsa is independent of timeSaccade Amplitude
Acquisition
# sticsa and iu group
t_test_independence_sticsa_iu_group_acq_sacc_amplitude <-
t.test(
df_long_acq_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "1", "sticsa_total_centred"],
df_long_acq_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_iu_group_acq_sacc_amplitude##
## Two Sample t-test
##
## data: df_long_acq_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "1", "sticsa_total_centred"] and df_long_acq_sacc_amplitude[df_long_acq_sacc_amplitude$iu_group == "-1", "sticsa_total_centred"]
## t = 9.3255, df = 276, p-value < 0.00000000000000022
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 7.343247 11.273157
## sample estimates:
## mean of x mean of y
## 4.754549 -4.553653# p < .05 : sticsa is not independent of iu group
# sticsa and stimulus
t_test_independence_sticsa_stimulus_acq_sacc_amplitude <-
t.test(
df_long_acq_sacc_amplitude[df_long_acq_sacc_amplitude$stimulus == "1", "sticsa_total_centred"],
df_long_acq_sacc_amplitude[df_long_acq_sacc_amplitude$stimulus == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_stimulus_acq_sacc_amplitude##
## Two Sample t-test
##
## data: df_long_acq_sacc_amplitude[df_long_acq_sacc_amplitude$stimulus == "1", "sticsa_total_centred"] and df_long_acq_sacc_amplitude[df_long_acq_sacc_amplitude$stimulus == "-1", "sticsa_total_centred"]
## t = 0, df = 276, p-value = 1
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -2.252832 2.252832
## sample estimates:
## mean of x mean of y
## -0.000000000000002862893 -0.000000000000002862893# p > .05 - sticsa is independent of stimulusExtinction
# sticsa and iu group
t_test_independence_sticsa_iu_group_ext_sacc_amplitude <-
t.test(
df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "1", "sticsa_total_centred"],
df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_iu_group_ext_sacc_amplitude##
## Two Sample t-test
##
## data: df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "1", "sticsa_total_centred"] and df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$iu_group == "-1", "sticsa_total_centred"]
## t = 13.212, df = 554, p-value < 0.00000000000000022
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## 7.924338 10.692067
## sample estimates:
## mean of x mean of y
## 4.754549 -4.553653# p < .05 : sticsa is not independent of iu group
# sticsa and stimulus
t_test_independence_sticsa_stimulus_ext_sacc_amplitude <-
t.test(
df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$stimulus == "1", "sticsa_total_centred"],
df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$stimulus == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_stimulus_ext_sacc_amplitude##
## Two Sample t-test
##
## data: df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$stimulus == "1", "sticsa_total_centred"] and df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$stimulus == "-1", "sticsa_total_centred"]
## t = 0, df = 554, p-value = 1
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -1.586608 1.586608
## sample estimates:
## mean of x mean of y
## -0.000000000000002862855 -0.000000000000002862855# p > .05 - sticsa is independent of stimulus
# sticsa and time
t_test_independence_sticsa_time_ext_sacc_amplitude <-
t.test(
df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$time == "1", "sticsa_total_centred"],
df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$time == "-1", "sticsa_total_centred"],
var.equal = TRUE
)
t_test_independence_sticsa_time_ext_sacc_amplitude##
## Two Sample t-test
##
## data: df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$time == "1", "sticsa_total_centred"] and df_long_ext_sacc_amplitude[df_long_ext_sacc_amplitude$time == "-1", "sticsa_total_centred"]
## t = 0, df = 554, p-value = 1
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
## -1.586608 1.586608
## sample estimates:
## mean of x mean of y
## -0.000000000000002862855 -0.000000000000002862855# p > .05 - sticsa is independent of timeHomogeneity of Regression Slopes
####### check homogeneity of regression slopes
df_long_acq_fix_count %>%
anova_test(fix_count ~ sticsa_total_centred + iu_group + stimulus + iu_group*stimulus +
sticsa_total_centred*iu_group + sticsa_total_centred*stimulus +
sticsa_total_centred*iu_group*stimulus)## Coefficient covariances computed by hccm()## ANOVA Table (type II tests)
##
## Effect DFn DFd F p p<.05 ges
## 1 sticsa_total_centred 1 270 0.114 0.736 0.0004210
## 2 iu_group 1 270 6.146 0.014 * 0.0220000
## 3 stimulus 1 270 0.957 0.329 0.0040000
## 4 iu_group:stimulus 1 270 0.103 0.749 0.0003810
## 5 sticsa_total_centred:iu_group 1 270 3.336 0.069 0.0120000
## 6 sticsa_total_centred:stimulus 1 270 0.154 0.695 0.0005710
## 7 sticsa_total_centred:iu_group:stimulus 1 270 0.021 0.885 0.0000783# as the interaction term sticsa_total_centred*iu_group was not sig (p = .067), there
# is not an interaction between iu group and sticsa total centred
# as the interaction term sticsa_total_centred*stimulus was not sig (p = .787), there
# is not an interaction between stimulus and sticsa total centred
# There was homogeneity of regression slopes as the interaction terms between
# the covariate (STICSA total centred) and grouping variables (iu group
# and stimulus), was not statistically significant, p > .05.
# reporting example:
# There was homogeneity of regression slopes as the interaction term was not statistically significant, F(2, 39) = 0.13, p = 0.88.
######## check remaining assumptions
# to check remaining assumptions, first compute lm() version of the model
ancova_lm_model_acq_fix_count <- lm(fix_count ~ sticsa_total_centred + iu_group * stimulus, df_long_acq_fix_count)
model_metrics <- augment(ancova_lm_model_acq_fix_count)
shapiro_test(model_metrics$.resid)## # A tibble: 1 × 3
## variable statistic p.value
## <chr> <dbl> <dbl>
## 1 model_metrics$.resid 0.940 0.00000000340summary(ancova_lm_model_acq_fix_count)##
## Call:
## lm(formula = fix_count ~ sticsa_total_centred + iu_group * stimulus,
## data = df_long_acq_fix_count)
##
## Residuals:
## Min 1Q Median 3Q Max
## -5.4309 -2.6331 -0.4267 2.0982 15.4692
##
## Coefficients:
## Estimate Std. Error t value Pr(>|t|)
## (Intercept) 6.706298 0.420831 15.936 <0.0000000000000002 ***
## sticsa_total_centred 0.008328 0.024721 0.337 0.7365
## iu_group1 1.224693 0.623745 1.963 0.0506 .
## stimulus1 -0.341613 0.573457 -0.596 0.5519
## iu_group1:stimulus1 -0.120642 0.819886 -0.147 0.8831
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## Residual standard error: 3.417 on 273 degrees of freedom
## Multiple R-squared: 0.03625, Adjusted R-squared: 0.02213
## F-statistic: 2.567 on 4 and 273 DF, p-value: 0.03851# obtain the assumptions and save as variable
q1_ancova_assumptions <- check_model(ancova_lm_model_acq_fix_count)
# check the assumptions
q1_ancova_assumptions## Registered S3 methods overwritten by 'parameters':
## method from
## as.double.parameters_kurtosis datawizard
## as.double.parameters_skewness datawizard
## as.double.parameters_smoothness datawizard
## as.numeric.parameters_kurtosis datawizard
## as.numeric.parameters_skewness datawizard
## as.numeric.parameters_smoothness datawizard
## print.parameters_distribution datawizard
## print.parameters_kurtosis datawizard
## print.parameters_skewness datawizard
## summary.parameters_kurtosis datawizard
## summary.parameters_skewness datawizard## Warning: `guides(<scale> = FALSE)` is deprecated. Please use `guides(<scale> =
## "none")` instead.## Loading required namespace: qqplotr
######## normality of residuals
# panel 2: most data points are roughly plotted along the line, but there do appear
# to be some possible deviations to assumption of normality
# panel 3: residuals appear mostly normally distributed,
# though there are some peaks on both tails
# assumption violated
###### homoscedasticity
# panel 4: data-points roughly plotted along horizontal line
# assumption is met
####### outliers
# panel 6: no cases exceed Cook's value of 1, so no outliers by influence.
# assumption is met.Outliers
## acquisition fix count
df_long_acq_fix_count %>%
group_by(iu_group, stimulus) %>%
identify_outliers(fix_count)## # A tibble: 4 × 10
## iu_group stimulus id sticsa_total condition fix_count sticsa_total_avg
## <fct> <fct> <fct> <dbl> <fct> <dbl> <dbl>
## 1 1 -1 086_1 68 acq_csm_fix_c… 18.3 40.5
## 2 1 -1 099_1 52 acq_csm_fix_c… 16 40.5
## 3 1 1 086_1 68 acq_csp_fix_c… 23.2 40.5
## 4 1 1 099_1 52 acq_csp_fix_c… 20.2 40.5
## # … with 3 more variables: sticsa_total_centred <dbl>, is.outlier <lgl>,
## # is.extreme <lgl>## extinction fix count
df_long_ext_fix_count %>%
group_by(iu_group, stimulus, time) %>%
identify_outliers(fix_count)## # A tibble: 14 × 11
## iu_group stimulus time id sticsa_total condition fix_count
## <fct> <fct> <fct> <fct> <dbl> <fct> <dbl>
## 1 -1 -1 -1 122_1 37 l_ext_csm_fix_count 17.8
## 2 -1 -1 1 047_1 41 e_ext_csm_fix_count 17
## 3 -1 -1 1 122_1 37 e_ext_csm_fix_count 21.5
## 4 -1 1 -1 122_1 37 l_ext_csp_fix_count 16
## 5 -1 1 1 122_1 37 e_ext_csp_fix_count 20.5
## 6 -1 1 1 143_1 44 e_ext_csp_fix_count 19.5
## 7 1 -1 -1 033_1 54 l_ext_csm_fix_count 20
## 8 1 -1 -1 065_1 33 l_ext_csm_fix_count 14.8
## 9 1 -1 -1 086_1 68 l_ext_csm_fix_count 19.2
## 10 1 -1 -1 099_1 52 l_ext_csm_fix_count 16
## 11 1 -1 -1 113_1 31 l_ext_csm_fix_count 15
## 12 1 1 -1 086_1 68 l_ext_csp_fix_count 22
## 13 1 1 1 086_1 68 e_ext_csp_fix_count 19.2
## 14 1 1 1 113_1 31 e_ext_csp_fix_count 17.8
## # … with 4 more variables: sticsa_total_avg <dbl>, sticsa_total_centred <dbl>,
## # is.outlier <lgl>, is.extreme <lgl># acquisition fix duration log
df_long_acq_fix_duration_log %>%
group_by(iu_group, stimulus) %>%
identify_outliers(fix_duration_log)## [1] iu_group stimulus id
## [4] sticsa_total condition fix_duration_log
## [7] sticsa_total_avg sticsa_total_centred is.outlier
## [10] is.extreme
## <0 rows> (or 0-length row.names)## extinction fix count
df_long_ext_fix_duration_log %>%
group_by(iu_group, stimulus, time) %>%
identify_outliers(fix_duration_log)## # A tibble: 6 × 11
## iu_group stimulus time id sticsa_total condition fix_duration_log
## <fct> <fct> <fct> <fct> <dbl> <fct> <dbl>
## 1 1 -1 -1 009_1 41 l_ext_csm_fix_dur… 8.00
## 2 1 -1 -1 010_1 43 l_ext_csm_fix_dur… 4.69
## 3 1 -1 -1 015_1 55 l_ext_csm_fix_dur… 8.37
## 4 1 -1 -1 044_1 36 l_ext_csm_fix_dur… 7.91
## 5 1 -1 1 044_1 36 e_ext_csm_fix_dur… 8.70
## 6 1 -1 1 113_1 31 e_ext_csm_fix_dur… 4.18
## # … with 4 more variables: sticsa_total_avg <dbl>, sticsa_total_centred <dbl>,
## # is.outlier <lgl>, is.extreme <lgl>## acquisition sacc amplitude
df_long_acq_sacc_amplitude %>%
group_by(iu_group, stimulus) %>%
identify_outliers(sacc_amplitude)## # A tibble: 9 × 10
## iu_group stimulus id sticsa_total condition sacc_amplitude sticsa_total_avg
## <fct> <fct> <fct> <dbl> <fct> <dbl> <dbl>
## 1 -1 -1 016_1 26 acq_csm_… 7.37 40.5
## 2 -1 -1 026_1 55 acq_csm_… 6.87 40.5
## 3 -1 1 016_1 26 acq_csp_… 6.35 40.5
## 4 1 -1 017_1 33 acq_csm_… 7.81 40.5
## 5 1 -1 021_1 54 acq_csm_… 7.50 40.5
## 6 1 -1 022_1 50 acq_csm_… 8.57 40.5
## 7 1 1 009_1 41 acq_csp_… 7.47 40.5
## 8 1 1 043_1 39 acq_csp_… 6.88 40.5
## 9 1 1 044_1 36 acq_csp_… 8.15 40.5
## # … with 3 more variables: sticsa_total_centred <dbl>, is.outlier <lgl>,
## # is.extreme <lgl>## extinction sacc amplitude
df_long_ext_sacc_amplitude %>%
group_by(iu_group, stimulus, time) %>%
identify_outliers(sacc_amplitude)## # A tibble: 17 × 11
## iu_group stimulus time id sticsa_total condition sacc_amplitude
## <fct> <fct> <fct> <fct> <dbl> <fct> <dbl>
## 1 -1 -1 -1 016_1 26 l_ext_csm_sacc_amp… 10.9
## 2 -1 -1 -1 075_1 35 l_ext_csm_sacc_amp… 8.98
## 3 -1 -1 -1 078_1 42 l_ext_csm_sacc_amp… 8.03
## 4 -1 -1 -1 111_1 41 l_ext_csm_sacc_amp… 8.21
## 5 -1 -1 1 016_1 26 e_ext_csm_sacc_amp… 9.11
## 6 -1 1 -1 016_1 26 l_ext_csp_sacc_amp… 13.1
## 7 -1 1 -1 075_1 35 l_ext_csp_sacc_amp… 7.84
## 8 -1 1 1 016_1 26 e_ext_csp_sacc_amp… 9.02
## 9 -1 1 1 051_1 28 e_ext_csp_sacc_amp… 7.40
## 10 -1 1 1 119_1 43 e_ext_csp_sacc_amp… 9.18
## 11 1 -1 -1 009_1 41 l_ext_csm_sacc_amp… 8.00
## 12 1 -1 -1 105_1 33 l_ext_csm_sacc_amp… 9.74
## 13 1 -1 1 105_1 33 e_ext_csm_sacc_amp… 11.4
## 14 1 1 -1 009_1 41 l_ext_csp_sacc_amp… 9.62
## 15 1 1 -1 022_1 50 l_ext_csp_sacc_amp… 8.34
## 16 1 1 1 009_1 41 e_ext_csp_sacc_amp… 7.67
## 17 1 1 1 129_1 46 e_ext_csp_sacc_amp… 8.65
## # … with 4 more variables: sticsa_total_avg <dbl>, sticsa_total_centred <dbl>,
## # is.outlier <lgl>, is.extreme <lgl>### reporting:
# There were no extreme outliers.